The management of advertising algorithms on the Internet. A case
study: Facebook and Google
La gestión de los algoritmos publicitarios en Internet. Un caso de estudio:
Facebook y Google
doxa.comunicación |
nº 36, pp. 243-271
| 243
January-June of 2023
ISSN: 1696-019X / e-ISSN: 2386-3978
How to cite this article:
Luque Ortiz, S. (2023). e management of advertising algorithms on the Internet. A case study: Facebook and Google.
Doxa Comunicación
,
36, pp. 243-271.
https://doi.org/10.31921/doxacom.n36a1713
Sergio Luque Ortiz.
PhD in Social Communication, professor at the International University of La Rioja (UNIR) and the
Miguel de Cervantes European University (UEMC) where he teaches in the faculties of Education and Social Sciences,
respectively, in the master’s degrees Teacher Training, Digital Marketing and SEO, as well as in the ocial degrees of Jour
-
nalism, ADE, Communication, Advertising and Public Relations. Sergio Luque researches dierent areas related to Social
Sciences such as Journalism, ethics of journalistic media, Digital Marketing, consumption of television and media audi
-
ences, as well as the rise of new digital communication formats. Collaborating researcher at the University of Verona (Italy),
member of the International Network of Researchers in Design at the University of Palermo (Buenos Aires, Argentina), as
well as a collaborator in dierent committees of scientic journals.
Universidad Europea Miguel de Cervantes, Spain
[email protected]
ORCID: 0000-0002-4302-9503
Abstract:
e birth of the Internet and the new information and communication
technologies have generated a new communicative paradigm. In
fact, it has also generated a new advertising paradigm with multiple
opportunities obtained through segmentation or the inclusion of
algorithms in search engines or social networks for product promotion.
Besides, there is the ability to measure the results achieved. In this
regard, this research addresses the management of advertising
algorithms executed in Facebook ADS and Google AdWords tools as
a new online advertising model compared to traditional approaches.
erefore, a qualitative and quantitative methodology has been
chosen to gather the information (using interviews and surveys) to
deepen into the dierences between the conventional advertising
management and online advertising. Among the results obtained, we
may advance that advertising has led to a new scenario where machine
learning will condition the future of advertising with the help of big
data and content programming.
Keywords:
Algorithms;
Internet; Google; Facebook; online advertising.
Resumen:
El nacimiento de Internet y las nuevas tecnologías de información y co
-
municación no solo han generado un nuevo paradigma comunicativo,
sino también publicitario con múltiples oportunidades obtenidas en
base a la segmentación, la promoción de productos mediante la inclu
-
sión de algoritmos en los buscadores y en las redes sociales, además de la
capacidad para medir los resultados logrados. Al respecto, esta investi
-
gación aborda la gestión de los algoritmos publicitarios ejecutada en las
herramientas Facebook ADS y Google Adwords como un nuevo modelo
de la publicidad online frente a los enfoques tradicionales. Para ello y,
con la nalidad de ahondar más sobre las diferencias entre el sistema de
gestión publicitaria convencional y la publicidad online, se ha optado
por una metodología tanto cualitativa como cuantitativa basada en la
utilización de la entrevista y la encuesta, respectivamente, como méto
-
dos de recolección de datos. Entre los resultados obtenidos puede avan
-
zarse que la publicidad ha derivado en uno nuevo escenario en el que el
machine learning condicionará el futuro de la publicidad ayudándose
del big data y la programación de contenidos.
Palabras clave:
Algoritmos; Internet; Google; Facebook; publicidad online.
Received: 07/06/2022 - Accepted: 11/11/2022 - Early access: 14/11/2022 - Published: 01/01/2023
Recibido: 07/06/2022 - Aceptado: 11/11/2022 - En edición:14/11/2022 - Publicado: 01/01/2023
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
244
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
1. Introduction and State of the Question
e Internet has meant an unprecedented transformation in both communication and advertising. e evolution has been
constant. Pérez (2020) points out that conventional “Above the Line” advertising has given way to a new understanding
advertising message dissemination by means of “Below the Line” supports via email marketing actions or social networks
advertising.
In this regard, Muela (2008) states that the Internet has contributed to democratize advertising by becoming an accessible
tool. On the one hand, advertising investment on the Internet by companies has grown “compared to television, press and
magazines that are falling, although its digital side is also growing “(Sanchez, 2019, 10). On the other hand, digital advertising
attractiveness lies in the inclusion of elements that are easily accessible to audiences. De Salas (2010) argues that both social
networks and Internet search engines have made possible for all types of companies and institutions to carry out advertising
campaigns without prior knowledge or huge budgets.
Accordingly, González (2014) argues that Facebook and Google manage millions of people’s data around the world by
controlling the large Internet advertising market through search and promotion algorithms conguration. is author inquires
about the value proposition provided by Facebook and Google trying to explain the reasons for their dierences with respect
to other technology companies such as Twitter and Yahoo!
For Boyd (2018), one of Facebook’s keys for success has resided in the inclusion of an algorithm that has been improving as this
social network average user has done so. e expert highlights that publications, whether organic or advertising, are directed
towards all the people’ accounts registered in this network, guaranteeing an immediate eect.
Likewise, it is relevant to describe what an algorithm is, in order to understand the study relevance with deeper appreciation.
is concept, according to the Royal Spanish Language Academy, refers to dierent mathematical calculations applied with
the purpose of nding a coherent solution in the presence of multiple problems framed in various contexts.
Peña (2018) explains that algorithms are sets of rules, applications and instructions that, systematically applied to an input
data package, with a specic format, serve to solve a wide cataloguing of problems, challenges and unknowns present in
today’s societies. In this sense, Steiner considers that:
An algorithm is, in essence, a set of data and instructions to be supercially applied in order to obtain a pre-established result.
e information enters a given algorithm and the possible answers coming out from it, will be stored in terms of user information
(2012, 64).
Saéz (2020) describes that the 21st century is characterized by the emergence of automated decisions, i.e., choices made or
based on the use of articial intelligence (AI) applications and algorithmic systems, which, with or without human intervention,
achieve decision-making eects on people.
Similarly, Alonso and Fernández (2021) highlight that the algorithm is one of the most notorious parameters existing
in the creation of contemporary culture by forming an indisputable part of the processes of organization, socialization,
communication and dissemination of information through the Web 2.0 space.
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
245
In this regard, computers calculate data faster than the human brain. Algorithms, according to the aforementioned author, are
used not only to collect data related to users’ identities, but to predict electoral results, behaviours or consumption habits as
well. erefore, algorithms are introduced in all social spheres and areas.
In relation to the above, Rodriguez (2021) states that algorithms are programmed systematic operations to calculate the
previous actions that may occur facing certain problems. is theorist highlights the existing link between machine learning
and algorithms considering that they develop specic actions aimed at programming technological platforms, and data
capture systems and programs.
On the one hand, Montells (2021) points out that Facebook uses several algorithms, being Edgerank one of the most important.
is element determines the publications displayed to a registered user based on the activity produced by him/her. erefore,
Edgerank is a set of calculations, formulas and improvements established by Facebook with the aim of establishing the type
of content viewed by each registered prole. For Grané (2021), algorithms constitute a structured, created and determined
set of specic steps that have produced an enormously relevant concept today. In this regard, algorithms are hidden in social
networks as Facebook, in such a way that algorithms like EdgeRank determine the contents, images or videos shown on
Facebook timeline.
On the other hand, Google also uses its own algorithm: Pagerank. Authors such as Barriola et al. (2016) argue that Larry Page
and Sergey Brin created this algorithm in 1998. Pagerank ranks websites relevance paying attention to the links generated by
a page. For example, if website A contains a link to web page B, Pagerank interprets that the content generated by B is relevant
for web A and therefore indexed. Once a rst approach to the theoretical concept of algorithm has been made, it is relevant to
justify the choice of the study object. According to Enberg (2019), Facebook and Google represent two of the most important
and fastest growing platforms in terms of advertising campaigns management on the Internet through algorithms inclusion.
For the aforementioned author, the algorithms used by both companies are fundamental to understand more the current
advertising scenario dominated by the presence of an advertising dynamic that is increasingly persuasive, hyper-segmented
and aimed at the Internet consumer. erefore, it is necessary to know how these companies
’
algorithms work, leading to a
new advertising concept far from conventionalisms.
In this regard, advertising algorithms could not be understood without digital consumers’ data management. is is what is
known as big data. Authors such as Franks (2012) recognize that big data is real and here to stay. Far from being perceived as
a passing or one-o fad, user data management (e-mails, postal addresses, telephone numbers or social prole addresses,
among other information sources) has become a complex reality.
Ortiz (2021) states that big data is a term used to conceptualize a large amount of data or a combination of them so that the
description, storage and processing of information derived from this data accumulation acquires a notorious importance.
Algorithms process daily decisions and data related to the use of technological platforms like Facebook and Google, among
others.
Companies such as Facebook and Google know how to use Internet users’ data to obtain a substantial prot. However, authors
like Zuazo (2018) highlight the existing opacity around algorithms management, knowledge and use by the aforementioned
246
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
companies. e expert highlights that few employees, managers or engineers share data on the algorithms’ actual functioning
inside and outside these companies.
Dermak (2017) adds that once the algorithm gets the precise and needed information, it is able to transform the data obtained
into a very powerful asset that technology sector companies are not willing to share. In relation to this, and once an adequate
theoretical denition of the study object of this research has been provided, it is necessary to describe the work’s objectives,
which are the following:
–
-Objective 1. To analyse advertising algorithms management in Facebook ADS and Google Adwords as they are two of the
advertising channels or supports most used by companies.
–
Objective 2. To describe the criteria used by algorithms, in addition to automated functionalities.
–
Objective 3. To address the existing dierences in advertising algorithms in terms of budget allocation, ad optimization
and audience detection.
Likewise, several starting hypotheses are raised within the present research, which are set out below:
–
Hypothesis 1. Facebook and Google algorithms perform better advertising campaign budget allocation as compared to
manual allocation. Algorithms predictive ability and impact are greater than human knowledge.
–
Hypothesis 2. Facebook and Google algorithms allow a higher level of created ads analysis based on measurement criteria,
impact and audience penetration that are dicult to detect by professionals running advertising campaigns.
–
Hypothesis 3. Facebook and Google’s advertising algorithms serve to reach audience groups that are related to products,
brands and services when compared to the manually reached segmentation.
–
Hypothesis 4. Facebook and Google’s advertising algorithms are one of the keys to the success of these platforms, which do
not clearly reveal how they work and handle the information, data and other elements compiled.
1.1. Online and programmatic advertising conceptualization
It is necessary to analyse the evolution of advertising to understand the boom and subsequent consolidation that both online
and programmatic advertising have experienced. e communication paradigm change has shaped a new reality where
consumers and brands converge with each other in multiple scenarios.
Alcalá (2021) states that, with the appearance of the Internet in 1989, there was a great change dealing with supply, demand
and advertising campaigns’ creation, with a progressive modernization process on the part of advertisers. For their part,
experts such as Papi-Gálvez et al. (2014) state that segmentation, personalization and participation are the three unique online
advertising distinguishing features, being the three potential attraction factors for many companies.
Tobias, Vallejo and Hinojo (2021) argue that online advertising has become one of the most important tools for companies
to reach consumers in today’s societies. e accumulation and exploitation of data obtained through digital programs oer a
wide range of possibilities for personalization, measurement and negotiation of Internet users’ information through multiple
options.
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
247
However, the multiplicity of models has not always been like that. Martínez, Segura and Sánchez (2011) do not hesitate to
conrm that the evolution of Internet advertising should be understood under a holistic approach including what is known as
Web 1.0, or the static 1990 web, Web 2.0, established at the beginning of the new millennium, and nally, semantic or Web 3.0.
Regarding advertising evolution at the dawn of the Internet and social networks, De Salas (2011) considers that online
advertising oers an added value based on the very act of selling an article or product, reinforcing not only the consumer’s
brand adhesion, but also the symbolic discourse implicit in any advertising action. Ramos (2006) describes that the current
online environment has served to promote the creation of a perfect symbiosis between advertising and entertainment whose
main purpose is to capture the public by using shocking, daring and original messages.
Similarly, online advertising is related to programmatic advertising by sharing origin and features. Rodriguez (2016) points out
that the advertising processes automation is a constant fact, with unstoppable growth and a high probability of becoming an
increasingly eective advertising investment model.
Villarreal (2022) states that programmatic advertising is a type of online advertising characterized by the automated, targeted
and segmented purchase of Internet advertising spaces that are chosen based on the audience and its characteristics.
Programmatic advertising connects brands with consumers in a clear, eective and fast way, generating a more direct link
between the two. In this way, a native or automated advertising strategy is generated.
e fast evolution of the Internet has led to the emergence of programmatic advertising design platforms. In 1994, the online
magazine HotWired, in collaboration with the American telecommunications company AT&T, was the rst company to
devise a sale model of digital advertising space, breaking into a new way of monetizing the expenditure and prot obtained in
advertising purchase and sale.
is gave rise to the cost per thousand impressions (CPT), a model where the advertiser only pays for the number of times
the ad is shown to users. Later, Facebook, through the Facebook Business Manager tool allowing the advertising campaigns
management, and Google, through Google AdWords, have established an auction and cost per click model (CPC, hereinafter)
determining a new way of advertising.
e union of CPT and CPC means the establishing of programmatic advertising, a buying and selling Internet spaces modality
that not only impacts among consumers, but also saves time and resources in what has to do with advertising management.
For Placebo Media (2016), the advertising space purchase through direct bidding (RTB, hereinafter) involves the use of
technology that allows buying, selling and collecting impressions under a real-time bidding paradigm, automatically achieving
impressions after impressions, users and reactions based on products, services and brands segmentations. In relation to the
above, Sevillano points out that:
RTB means the capacity to purchase advertising space in an automated way, access inventory and buy it in real time without
human intervention and programmatic advertising goes further in the sense that it starts from the parameter that the buying party
decision making chooses which advertiser and values how much it is willing to pay for a user, i.e., who (s)he is, their interests, the
user relevance for the brand (2015, 10).
Programmatic advertising could not be understood without Facebook Business Manager and Google AdWords. Costalago
(2019) conrms that, in Google’s case, the appearance of Google Adwords advertising tool in 2004 has meant the programmatic
248
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
advertising consolidation. For this author, Google Adwords has become the main system for buying Internet advertising space,
far ahead of other tools such as Facebook Business Manager.
e advertising campaigns management through Facebook Business Manager not only allows the selection of very specic
campaign aspects such as user locations, audience interests, impacts duration and advertising budgets, but also oers the
opportunity to synchronize the results achieved with other networks such as Instagram.
1.2. Big data, Articial Intelligence (AI) and Machine Learning
It is clear to understand the importance of capturing user data, i.e., big data. In this perspective, Puyol (2015) considers that
big data is the massive volume of data that adopts the information collected through the Internet. e intrinsic meaning of big
data points to the great opportunity oered by technological tools.
Other authors such as Chen et al. (2013) describe big data as an information asset of high volume, speed and variety that
requires cost-eective decision making and the application of novel data collection formulas. For Camargo-Vega, Camargo-
Ortega and Joyanes (2014) the term big data concentrates a series of characteristics such as the following:
–
Accumulation of data complex to analyse without specic tools.
–
Big data processing and analysis.
–
Application of procedures that allow the creation of big datasets.
–
Information analysis and processing to extract novel ideas.
–
Technological development that makes the use of big data more economical.
–
Consolidation of a company’s information to make it available to the public.
Raya (2015) considers that the information currently managed by companies through data capture on the Internet must
guarantee the users’ non-discrimination. For this purpose, various algorithms are used to ascertain a unied data treatment.
Estera (2016) conrms that big data constitutes a fundamental part of the strategies created by both companies and the State
with the aim of processing data more eciently.
Similarly, big data could hardly be understood without AI, a concept in vogue in recent years but which is not recent. Marvin
Mansky was one of the rst theorists to outline a denition of this term in 1968. Manksy (1968) conrmed that AI origin should
be understood under the gure of Alan Turing, one of the rst mathematical algebra pioneers. For Peña (2010), there were
various problems shown to require the use of technology to nd solutions adapted to the challenges created from about the
middle of last century.
Both Google and Facebook employ AI with the aim of improving user experience, advertising campaigns investment and web
interface. Han (2014) does not hesitate to state that both companies’ algorithms analyse the signals sent by users every time
they browse their platforms.
Predicting behaviour and consumption habits of Internet users is one of the most important purposes of programmatic
advertising. For this purpose, Facebook and Google use AI as a very useful tool. In the opinion of Vallverdú (2019), algorithms’
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
249
use should not be understood without the usefulness provided by articial intelligence (AI) in the dicult and complex task
of anticipating consumer needs.
It is complex to understand how the advertising algorithms used by the companies studied in this article work. For this reason,
it is necessary to analyse what the concept of machine learning involves. For Mitchell (1997) this term describes automatic
learning. e essence of machine learning consists of the processing and selection of information obtained through big data
application. Algorithms make predictions about Internet users’ behaviour thanks to machine learning, which for Gupta (2017),
is fundamental in online and programmatic advertising.
Jung (2022) states that machine learning aims to predict or guess the features that a dataset presents. Any machine learning
method must be based on the implementation of innite computational resources to obtain the largest amount of informative
data.
In light of the above, authors such as Martínez, Aguado and Sánchez (2022) describe that AI impact on advertising is relevant for
three reasons. e rst involves the advertising business model transformation because of the Internet and online advertising
platforms like Google AdWords. e second one refers to the technological change experienced in advertising and digital
native industries. e third and last one has to do with AI strategic interest in the digital societies of the 21st century.
However, programmatic advertising managed through strategies including algorithms and AI leaves behind multiple criticisms.
Iniesta-Alemán et al. (2018) claim that more conventional advertising agencies may eventually disappear because advertisers
will be autonomous in campaign management.
2. Methodology
e present work is characterized by a qualitative and quantitative methodological approach based on the use of a structured
interview and a survey, respectively. In view of the above, Vargas (2012) describes that the structured interview is dened by
the design of a series of questions to be answered based on a predetermined limit of responses. erefore, this type of interview
is elaborated in advance and is posed to the participants with rigidity and systematization.
Lucca and Berríos (2003) consider that this type of interview is very formal, allowing the interviewer to compare the information
obtained based on the answers given by the interviewees. Del Rincón, Arnal, Latorre and Sans (1995) say that in the structured
interview the interviewer asks each of the interviewees dierent questions with a prior limitation of answers. In this regard,
the aforementioned authors emphasize that a protocol of prexed questions and answers is elaborated and must be followed
with rigidity and no improvisation.
For their part, Denzin and Lincoln (2005) argue that the researcher carries out a prior planning of all the questions he wants
to ask. To do so, he/she must prepare a script with the questions sequenced and ordered. Similarly, the interviewee may not
make comments, value judgments or appreciations. e questions have a very specic nature: they are closed.
is implies that the answers can be armative, negative or specic on a series of possible answer options. In addition to what
has been described, Diaz-Bravo, Torruco-García, Martínez-Hernández and Varela-Ruiz (2013) conrm that the structured
interview must include a set of categories or possible options for the subject to choose from. e interview has to be applied
250
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
rigidly to all research participants, so that the order in which the questions are asked and the answers obtained are always the
same.
After the above, other aspects related to the present research are presented. First, an initial phase based on the execution
of a bibliographic study was carried out to contextualize the present study object from the perspective of contemporaneity.
Secondly, a survey was designed with 10 questions that included an evaluation method using ve variants of responses ranging
from 1 (totally disagree) to 5 (totally agree).
e sample is composed of three hundred professionals (in dierent locations in Europe, Asia and USA) related to programmatic
advertising, advertising algorithms management on Facebook and Google and user database management. e number of
participants has been established at three hundred (one hundred for each geographical area) due to the specic nature of
the sample, i.e. professionals who accredit having the Certicate in Data Analytics Professional (known as CDAP) and the
Professional Certicate in Data Analytics (known as PCDA) issued by Google Global. e contact with the participants was
executed through a previous stage of proles search and selection carried out by the researcher.
For reasons of compliance with the current European regulations of the Organic Law on Data Protection (OLDP, hereinafter),
the specic personal data of each of the interviewees have been concealed, guaranteeing both anonymity and the correct
treatment of the recapitulated information during the research course. Due to the geographical disparity of the respondents,
the data collection process was carried out online. All survey participants took part in the survey with no unanswered questions
or blank items. In this regard, all participants of the present study received an e-mail with a link containing the survey. e
survey template provided is shown below.
Table 1. Survey model designed in the research
Questions
Value 1Value 2
Value3
Value 4Value 5
1. Machine learning and the functionalities of Facebook
and Google’s advertising algorithms are fundamental to
my online strategy.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
2. e inclusion of automated actions through
programmatic advertising on Facebook and Google is
essential to improve campaign results.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
3. e automatic allocation of budgets in advertising
campaigns on Facebook and Google enables better
results than manual allocation.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
4. e automatic allocation of budgets in advertising
campaigns allows me to save time and resources in terms
of advertising management.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
251
5. Facebook and Google’s algorithm allows me to reach
a group of like-minded audiences more accurately than
manual segmentation.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
6. Facebook and Google’s advertising algorithms
are more ecient in detecting creatives improving
conversions compared to manual management of
creatives.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
7. e future of online and programmatic advertising
depends on the knowledge of Facebook and Google’s
advertising interfaces.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
8. Digital advertising campaigns increasingly require the
use of algorithms and automation that Facebook and
Google oer.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
9. e evolution of machine learning will increasingly
include actions performed by algorithms as opposed to
services provided by human capital.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
10. e handling of information from big data is crucial
to understand the proper functioning of advertising
algorithms.
Totally
disagree
Somewhat
disagree
Neither agree
nor disagree
Somewhat
agree
Totally agree
Source: prepared by the author
Similarly, the interview was used as a data collection tool. For this purpose, a sample of three hundred proles has been
designed, these being dierent from those participating in the survey. is is intended to guarantee not only the greatest
possible data plurality, but also to ensure the information contrast and its objectivity. Accordingly, the interview participants’
selection was carried out following the same requirements and patterns as with the survey. e interviewees prole has been
dened in three typologies dened below.
1.
Executives (from technology sector companies) who manage online and programmatic advertising teams that design
advertising campaigns.
2.
Internet advertising sector professionals, with proven experience in campaign implementation through algorithm
management.
3.
Facebook and Google partners with extensive knowledge on both algorithms and platforms operation.
Likewise, and in compliance with the legal precepts contained in the current OLDP, the anonymity of the interviewees is
guaranteed. e interview model is attached below.
252
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Table 2. Interview model
1. What are the competitive advantages of including automated functionalities by Facebook and Google?
-It helps in what is involved with creative process and decision-making.
-It simplies tasks, saves time and resources.
-It boosts campaign results.
2. What areas generate some distrust regarding advertising algorithms eectiveness?
-Human hand substitution vs. technology inclusion
-Use complexity
-Diculty in establishing a decalogue or good practices guide
3. Do you consider that both machine learning and advertising processes automation will result in the reduced inclusion of specialized person
-
nel in advertising campaign management?
-No. e inclusion of human resources will continue to be an essential asset.
-Yes. Algorithmic technology will undoubtedly play an increasingly important role.
-It will depend on each company, but machines will not be able to overtake humans.
4. Is it dangerous or controversial to automate so many functions by including advertising algorithms in platforms like Facebook and Google?
-No, because it is necessary to perform a weighted assignment of tasks.
-Yes, because more and more actions are performed automatically.
-ere is no clear or evident risk at the moment.
5. Do you think that the trend of investing more and more in Facebook and Google advertising will continue in the coming years or will com
-
panies return to conventional advertising campaigns?
-Yes. is will continue to be the case and will even increase.
-No. More and more companies are choosing hybrid campaigns.
-It will depend on each campaign, budget and advertising management.
6. Being a Facebook and Google partner, why do you think both platforms are so opaque when it comes to how the algorithms work?
-e lack of transparency has to do in part with advertising success.
-Because if algorithms were disclosed, they would cease to exist.
-e main reason for this is that advertising management generates very high prots.
7. Could Facebook and Google get more advertisers if they were more transparent about advertising algorithms management?
-Yes, especially for SMEs and companies with little knowledge of online and programmatic advertising.
-No, because Facebook and Google continue to reach advertisers and will do so regardless of the information that they release to the outside.
-Possibly, if this were the case, more and more companies would better understand why to invest in online advertising.
8. What characteristics should online and programmatic advertising experts and campaign managers meet?
-Algorithms require specic knowledge, so new job proles will appear.
-Online and programmatic advertising requires competitive proles that adapt to the new advertising and technological realities as op
-
posed to more conservative proles.
-New positions and job opportunities will emerge for the most active proles.
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
253
9. Could conventional advertising agencies disappear due to online advertising growth, and its multiple advantages such as the ability to gen
-
erate little budget advertising?
-Traditional advertising agencies will have to oer additional value to avoid disappearance or becoming Facebook and Google net
-
work add-ons.
-More conventional agencies must clearly distinguish themselves from other agency types growing in strength.
-ey will not disappear, but they will be reduced.
10. What would be your recommendation, as an online advertising professional expert, for all proles specializing in programmatic advertising
and advertising algorithms management?
-To be trained through master’s programs, specic courses and contents.
-To be self-taught by reading manuals, attending conferences...
-To keep a continuous training attitude in a changing scenario.
Source: prepared by the author
4. Results
Survey results
Chart 1. Machine learning and the functionalities of Facebook and Google’s advertising algorithms are fundamental in the online
strategy I carry out
12
-
To keep a continuous training attitude in a changing scenario.
Source:
p
repared by the author
4.
Results
-
Survey results.
C
hart 1. Machine learning and the functionalities of Facebook and Google's advertising
algorithms are fundamental in the online strategy I carry out
Source: prepared by the author
Sixty-
four point five percent
of
the participants stated that they totally agreed with the importance
that both Facebook's and Google's advertising algorithms have on the management of online
advertising campaigns, compared to 28.9% who somewhat agreed with this statement. A figure
of
3.
9% clearly indicated that they neither agreed nor disagreed, and finally 2.6% of the
respondents who stated that they totally disagreed.
Chart 2. The inclusion of automated actions through programmatic advertising on
Facebook and Google is essential to improve campaign results
Source: prepared by the author
3%
2%
10%
20%
65%
dŽƚĂůůLJ ĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Ϯй
ϭй
ϭϳй
ϰϮй
ϯϴй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
Sixty-four point ve percent of the participants stated that they totally agreed with the importance that both Facebook’s and
Google’s advertising algorithms have on the management of online advertising campaigns, compared to 28.9% who somewhat
agreed with this statement. A gure of 3.9% clearly indicated that they neither agreed nor disagreed, and nally 2.6% of the
respondents who stated that they totally disagreed.
254
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Chart 2. e inclusion of automated actions through programmatic advertising on Facebook and Google is essential to improve
campaign results
12
-
To keep a continuous training attitude in a changing scenario.
Source:
p
repared by the author
4.
Results
-
Survey results.
C
hart 1. Machine learning and the functionalities of Facebook and Google's advertising
algorithms are fundamental in the online strategy I carry out
Source: prepared by the author
Sixty-
four point five percent
of
the participants stated that they totally agreed with the importance
that both Facebook's and Google's advertising algorithms have on the management of online
advertising campaigns, compared to 28.9% who somewhat agreed with this statement. A figure
of
3.
9% clearly indicated that they neither agreed nor disagreed, and finally 2.6% of the
respondents who stated that they totally disagreed.
Chart 2. The inclusion of automated actions through programmatic advertising on
Facebook and Google is essential to improve campaign results
Source: prepared by the author
3%
2%
10%
20%
65%
dŽƚĂůůLJ ĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Ϯй
ϭй
ϭϳй
ϰϮй
ϯϴй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
irty-eight point two percent of the respondents did not hesitate to consider that the inclusion of automated actions in
programmatic advertising is appropriate for achieving better results. Forty-two point one percent agreed with this statement.
Similarly, Seventeen point one percent of respondents neither agreed nor disagreed, and nally 1.3% disagreed.
Chart 3. Automatic allocation of budgets in advertising campaigns on Facebook and Google leads to better results than manual allocation
13
Thirty-
eight point two percent of the respondents did not hesitate to consider that the inclusion of
automated actions in programmatic advertising is appropriate for achieving better results.
Forty-
two
point one percent agreed with this statement. Similarly, Seventeen point one percent of
respondents neither agreed nor disagreed, and finally 1.3% disagreed
.
Chart 3. Automatic allocation of budgets in advertising campaigns on Facebook and
Google leads to better results than manual allocation
Source: prepared by the author
Forty point eight percent
of
the participants stated that they agreed with the statement made in
question 3, compared to 31.6% who stated that they completely agreed. Compared to these
percentages, 14.5% disagreed with the statement, stating that manual assignment is more relevant
th
an automatic assignment. Similarly, 2.6% of the participants totally disagreed and 10.5% of
those surveyed said they neither agreed nor disagreed
.
Chart 4. The automatic allocation of budgets in advertising campaigns allows me to save
time and resources in terms of advertising management
Source: prepared by the author
A total of 44.7% of the respondents agreed with the time reduction implied by budget automation,
compared to 18.4% who totally agreed and another 18.4% who neither agreed nor disagreed on
ϱ͕ϬϬй
ϭϬй
ϭϬй
ϰϱй
ϯϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
ϱй
ϭϬй
ϮϬй
ϰϱй
ϮϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
255
Forty point eight percent of the participants stated that they agreed with the statement made in question 3, compared to
31.6% who stated that they completely agreed. Compared to these percentages, 14.5% disagreed with the statement, stating
that manual assignment is more relevant than automatic assignment. Similarly, 2.6% of the participants totally disagreed and
10.5% of those surveyed said they neither agreed nor disagreed.
Chart 4. e automatic allocation of budgets in advertising campaigns allows me to save time and resources in
terms of advertising management
13
Thirty-
eight point two percent of the respondents did not hesitate to consider that the inclusion of
automated actions in programmatic advertising is appropriate for achieving better results.
Forty-
two
point one percent agreed with this statement. Similarly, Seventeen point one percent of
respondents neither agreed nor disagreed, and finally 1.3% disagreed
.
Chart 3. Automatic allocation of budgets in advertising campaigns on Facebook and
Google leads to better results than manual allocation
Source: prepared by the author
Forty point eight percent
of
the participants stated that they agreed with the statement made in
question 3, compared to 31.6% who stated that they completely agreed. Compared to these
percentages, 14.5% disagreed with the statement, stating that manual assignment is more relevant
th
an automatic assignment. Similarly, 2.6% of the participants totally disagreed and 10.5% of
those surveyed said they neither agreed nor disagreed
.
Chart 4. The automatic allocation of budgets in advertising campaigns allows me to save
time and resources in terms of advertising management
Source: prepared by the author
A total of 44.7% of the respondents agreed with the time reduction implied by budget automation,
compared to 18.4% who totally agreed and another 18.4% who neither agreed nor disagreed on
ϱ͕ϬϬй
ϭϬй
ϭϬй
ϰϱй
ϯϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
ϱй
ϭϬй
ϮϬй
ϰϱй
ϮϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
A total of 44.7% of the respondents agreed with the time reduction implied by budget automation, compared to 18.4% who
totally agreed and another 18.4% who neither agreed nor disagreed on this issue. Similarly, 15.8% of the participants somewhat
disagreed with this statement, compared to 2.6% who completely disagreed.
256
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Chart 5. Facebook and Google’s algorithm allows me to reach a group of related audiences more accurately than manual segmentation
14
this issue. Similarly, 15.8% of the participants somewhat disagreed with this statement, compared
to 2.6% who completely disagreed.
Chart 5. Facebook and Google's algorithm allows me to reach a group of related audiences
more accurately than manual segmentation
Source: prepared by the author
Sixty point five percent of respondents completely agreed with the statement describing that the
advertising algorithms used by Facebook and Google serve to reach a group with related audiences
more accurately than manual targeting. This is followed by 14.5% of participants who somewhat
agreed with this idea, compared to 13.2% of members who had no clear position either for or
against, 7.92% who said they somewhat disagreed and 3.9% who showed a strongly disagreed
position.
Chart 6. Facebook and Google ad algorithms are more efficient in detecting creatives
improving conversions compared to manual management of the
creatives
Source: prepared by the author
Fifty-
one point three percent of respondents somewhat agreed with the statement that Facebook
and Google's advertising algorithms are more efficient at detecting creatives and improving
conversions compared to manual or traditional creative management. In this regard, 30.3% of the
par
ticipants fully agreed with this idea, compared to 17.1% who show a more neutralized position,
ϭϬй
ϯй
ϭϮй
ϲϬй
ϭϱй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
"ŐƌĞĞ
ϯй
Ϯй
ϭϱй
ϱϬй
ϯϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
Sixty point ve percent of respondents completely agreed with the statement describing that the advertising algorithms used
by Facebook and Google serve to reach a group with related audiences more accurately than manual targeting. is is followed
by 14.5% of participants who somewhat agreed with this idea, compared to 13.2% of members who had no clear position either
for or against, 7.92% who said they somewhat disagreed and 3.9% who showed a strongly disagreed position.
Chart 6. Facebook and Google ad algorithms are more ecient in detecting creatives improving conversions compared to manual
management of the creatives
14
this issue. Similarly, 15.8% of the participants somewhat disagreed with this statement, compared
to 2.6% who completely disagreed.
Chart 5. Facebook and Google's algorithm allows me to reach a group of related audiences
more accurately than manual segmentation
Source: prepared by the author
Sixty point five percent of respondents completely agreed with the statement describing that the
advertising algorithms used by Facebook and Google serve to reach a group with related audiences
more accurately than manual targeting. This is followed by 14.5% of participants who somewhat
agreed with this idea, compared to 13.2% of members who had no clear position either for or
against, 7.92% who said they somewhat disagreed and 3.9% who showed a strongly disagreed
position.
Chart 6. Facebook and Google ad algorithms are more efficient in detecting creatives
improving conversions compared to manual management of the
creatives
Source: prepared by the author
Fifty-
one point three percent of respondents somewhat agreed with the statement that Facebook
and Google's advertising algorithms are more efficient at detecting creatives and improving
conversions compared to manual or traditional creative management. In this regard, 30.3% of the
par
ticipants fully agreed with this idea, compared to 17.1% who show a more neutralized position,
ϭϬй
ϯй
ϭϮй
ϲϬй
ϭϱй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
"ŐƌĞĞ
ϯй
Ϯй
ϭϱй
ϱϬй
ϯϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
257
Fifty-one point three percent of respondents somewhat agreed with the statement that Facebook and Google’s advertising
algorithms are more ecient at detecting creatives and improving conversions compared to manual or traditional creative
management. In this regard, 30.3% of the participants fully agreed with this idea, compared to 17.1% who show a more
neutralized position, 1.3% with a 100% negative or contrary view with this idea so they consider that algorithms do not improve
conversions.
Chart 7. e future of online and programmatic advertising requires knowledge of Facebook and Google’s advertising interfaces
15
1.3% with a 100% negative or contrary view with this idea so they consider that algorithms do
not improve conversions.
C
hart 7. The future of online and programmatic advertising requires knowledge of
Facebook and Google's advertising interfaces
Source: prepared by the author
Fifty-
seven point nine percent of respondents somewhat agreed with the statement in the question,
followed by 19.7% who neither agreed nor disagreed, 18.4% who strongly agreed, and a minority
of 1.3% and 2.6%, respectively, who disagreed or strongly disagreed.
C
hart 8. Digital advertising campaigns increasingly require the use of algorithms and
automation offered by Facebook and
Google
Source: prepared by the author
Forty-
eight point seven percent
of
respondents agreed with the statement in question nº 8 that
Facebook and Google advertising campaigns increasingly require automation and the inclusion of
algorithms. A similar opinion is held by 46% of respondents. A 2.6% of the participants have a
neutral position (neither agree nor disagree) and, finally, other amounts expressed in groups who
have a contrary position
.
C
hart 9. Machine learning evolution will increasingly include actions performed by
algorithms as opposed to services provided by human capital
3%
2%
ϮϬй
ϱϱй
ϮϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
ϲй
Ϯй
Ϯй
ϱϬй
ϰϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
Fifty-seven point nine percent of respondents somewhat agreed with the statement in the question, followed by 19.7% who
neither agreed nor disagreed, 18.4% who strongly agreed, and a minority of 1.3% and 2.6%, respectively, who disagreed or
strongly disagreed.
258
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Chart 8. Digital advertising campaigns increasingly require the use of algorithms and automation oered by Facebook and Google
15
1.3% with a 100% negative or contrary view with this idea so they consider that algorithms do
not improve conversions.
C
hart 7. The future of online and programmatic advertising requires knowledge of
Facebook and Google's advertising interfaces
Source: prepared by the author
Fifty-
seven point nine percent of respondents somewhat agreed with the statement in the question,
followed by 19.7% who neither agreed nor disagreed, 18.4% who strongly agreed, and a minority
of 1.3% and 2.6%, respectively, who disagreed or strongly disagreed.
C
hart 8. Digital advertising campaigns increasingly require the use of algorithms and
automation offered by Facebook and
Google
Source: prepared by the author
Forty-
eight point seven percent
of
respondents agreed with the statement in question nº 8 that
Facebook and Google advertising campaigns increasingly require automation and the inclusion of
algorithms. A similar opinion is held by 46% of respondents. A 2.6% of the participants have a
neutral position (neither agree nor disagree) and, finally, other amounts expressed in groups who
have a contrary position
.
C
hart 9. Machine learning evolution will increasingly include actions performed by
algorithms as opposed to services provided by human capital
3%
2%
ϮϬй
ϱϱй
ϮϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
ϲй
Ϯй
Ϯй
ϱϬй
ϰϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
Forty-eight point seven percent of respondents agreed with the statement in question nº 8 that Facebook and Google advertising
campaigns increasingly require automation and the inclusion of algorithms. A similar opinion is held by 46% of respondents.
A 2.6% of the participants have a neutral position (neither agree nor disagree) and, nally, other amounts expressed in groups
who have a contrary position.
Chart 9. Machine learning evolution will increasingly include actions performed by algorithms as opposed to
services provided by human capital
16
Source:
P
repared by the author
This pie chart shows information on machine learning evolution and how this change has resulted
in more and more advertising actions being carried out by algorithms and machines supplanting
human identity. In this regard, 53.9% of the participants considered that they totally agreed,
followed by 39.5% who somewhat agreed, 2.6% who neither agreed nor disagreed, and finally,
3.9% who totally disagreed.
C
hart 10. The handling of big data information is key to understand the proper
advertising algorithms func
tioning
Source: prepared by the author
The survey ends with the tenth question, which asks whether big data information management is
so important to understand how advertising algorithms work. Faced with this situation, 40.8% of
the participants somewhat agreed, followed by 31.6% who completely agreed, 15.2% who said
they neither agreed nor disagreed, 11.8% who expressed some disagreement and 2.6% who totally
disagreed.
-
Results of the structured interview
C
hart 11. What are the competitive advantages of including automated functionalities by
Ϯй
ϯй
ϭϬй
ϯϱй
ϱϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
ϱй
ϭϬй
ϭϱй
ϰϬй
ϯϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: Prepared by the author
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
259
is pie chart shows information on machine learning evolution and how this change has resulted in more and more advertising
actions being carried out by algorithms and machines supplanting human identity. In this regard, 53.9% of the participants
considered that they totally agreed, followed by 39.5% who somewhat agreed, 2.6% who neither agreed nor disagreed, and
nally, 3.9% who totally disagreed.
Chart 10. e handling of big data information is key to understand the proper advertising algorithms functioning
16
Source:
P
repared by the author
This pie chart shows information on machine learning evolution and how this change has resulted
in more and more advertising actions being carried out by algorithms and machines supplanting
human identity. In this regard, 53.9% of the participants considered that they totally agreed,
followed by 39.5% who somewhat agreed, 2.6% who neither agreed nor disagreed, and finally,
3.9% who totally disagreed.
C
hart 10. The handling of big data information is key to understand the proper
advertising algorithms func
tioning
Source: prepared by the author
The survey ends with the tenth question, which asks whether big data information management is
so important to understand how advertising algorithms work. Faced with this situation, 40.8% of
the participants somewhat agreed, followed by 31.6% who completely agreed, 15.2% who said
they neither agreed nor disagreed, 11.8% who expressed some disagreement and 2.6% who totally
disagreed.
-
Results of the structured interview
C
hart 11. What are the competitive advantages of including automated functionalities by
Ϯй
ϯй
ϭϬй
ϯϱй
ϱϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
ϱй
ϭϬй
ϭϱй
ϰϬй
ϯϬй
dŽƚĂůůLJ ĚŝƐĂŐƌĞĞ
#ŝƐĂŐƌĞĞ
EĞŝƚŚĞƌ ĂŐƌĞĞ Žƌ ĚŝƐĂŐƌĞĞ
"ŐƌĞĞ
dŽƚĂůůLJ ĂŐƌĞĞ
Source: prepared by the author
e survey ends with the tenth question, which asks whether big data information management is so important to understand
how advertising algorithms work. Faced with this situation, 40.8% of the participants somewhat agreed, followed by 31.6% who
completely agreed, 15.2% who said they neither agreed nor disagreed, 11.8% who expressed some disagreement and 2.6% who
totally disagreed.
260
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Results of the structured interview
Chart 11. What are the competitive advantages of including automated functionalities by Facebook and Google?
17
Facebook and Google?
S
ource:
P
repared by the author
We can see that 40.60% of respondents consider that among the most obvious competitive
advantages of Facebook and Google automated functionalities are task simplification and time
a
nd resources savings. This is followed by 35% of respondents, compared to 24.40% who indicate
the relevance of boosting campaign results.
C
hart 12. Which areas generate some distrust regarding advertising algorithms
effectiveness?
Source:
P
repared by the author
Forty-
five percent of respondents indicated that the complexity involved in algorithm use is one
of the aspects that generates some mistrust. After this large percentage, 30.40% of the interviewees
point out the absence of a practice guide as one of the most important problems. Finally, 25% of
the interviewees point out human hand substitution against technology inclusion
.
Chart 13. Do you consider that both machine learning and advertising processes
automation will result in the reduced inclusion of specialized personnel in advertising
campaign management
?
ϯϱй
ϰϭй
Ϯϰй
"ƌĞĂƚŝǀĞ ƉƌŽĐĞƐƐ ĂŶĚ
ĚĞĐŝƐŝŽŶͲŵĂŬŝŶŐ ĂƐƐŝƐƚĂŶĐĞ
^ŝŵƉůŝĨŝĞƐ ƚĂƐŬƐ͕ ƐĂǀĞƐ ƚŝŵĞ
ĂŶĚ ƌĞƐŽƵƌĐĞƐ
!ŽŽƐƚƐ ĐĂŵƉĂŝŐŶ ƌĞƐƵůƚƐ
Ϯϱй
ϯϬй
ϰϱй
,ƵŵĂŶ ŚĂŶĚ ƐƵďƐƚŝƚƵƚŝŽŶ ǀƐ͘
ƚĞĐŚŶŽůŽŐLJ ŝŶĐůƵƐŝŽŶ
$ŝĨĨŝĐƵůƚLJ ŝŶ ĞƐƚĂďůŝƐŚŝŶŐ Ă
ĚĞĐĂůŽŐƵĞ Žƌ ŐŽŽĚ ƉƌĂĐƚŝĐĞƐ ŐƵŝĚĞ
hƐĞ ĐŽŵƉůĞdžŝƚLJ
Source: Prepared by the author
We can see that 40.60% of respondents consider that among the most obvious competitive advantages of Facebook and Google
automated functionalities are task simplication and time and resources savings. is is followed by 35% of respondents,
compared to 24.40% who indicate the relevance of boosting campaign results.
Chart 12. Which areas generate some distrust regarding advertising algorithms eectiveness?
17
Facebook and Google?
S
ource:
P
repared by the author
We can see that 40.60% of respondents consider that among the most obvious competitive
advantages of Facebook and Google automated functionalities are task simplification and time
a
nd resources savings. This is followed by 35% of respondents, compared to 24.40% who indicate
the relevance of boosting campaign results.
C
hart 12. Which areas generate some distrust regarding advertising algorithms
effectiveness?
Source:
P
repared by the author
Forty-
five percent of respondents indicated that the complexity involved in algorithm use is one
of the aspects that generates some mistrust. After this large percentage, 30.40% of the interviewees
point out the absence of a practice guide as one of the most important problems. Finally, 25% of
the interviewees point out human hand substitution against technology inclusion
.
Chart 13. Do you consider that both machine learning and advertising processes
automation will result in the reduced inclusion of specialized personnel in advertising
campaign management
?
ϯϱй
ϰϭй
Ϯϰй
"ƌĞĂƚŝǀĞ ƉƌŽĐĞƐƐ ĂŶĚ
ĚĞĐŝƐŝŽŶͲŵĂŬŝŶŐ ĂƐƐŝƐƚĂŶĐĞ
^ŝŵƉůŝĨŝĞƐ ƚĂƐŬƐ͕ ƐĂǀĞƐ ƚŝŵĞ
ĂŶĚ ƌĞƐŽƵƌĐĞƐ
!ŽŽƐƚƐ ĐĂŵƉĂŝŐŶ ƌĞƐƵůƚƐ
Ϯϱй
ϯϬй
ϰϱй
,ƵŵĂŶ ŚĂŶĚ ƐƵďƐƚŝƚƵƚŝŽŶ ǀƐ͘
ƚĞĐŚŶŽůŽŐLJ ŝŶĐůƵƐŝŽŶ
$ŝĨĨŝĐƵůƚLJ ŝŶ ĞƐƚĂďůŝƐŚŝŶŐ Ă
ĚĞĐĂůŽŐƵĞ Žƌ ŐŽŽĚ ƉƌĂĐƚŝĐĞƐ ŐƵŝĚĞ
hƐĞ ĐŽŵƉůĞdžŝƚLJ
Source: Prepared by the author
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
261
Forty-ve percent of respondents indicated that the complexity involved in algorithm use is one of the aspects that generates
some mistrust. After this large percentage, 30.40% of the interviewees point out the absence of a practice guide as one of the
most important problems. Finally, 25% of the interviewees point out human hand substitution against technology inclusion.
Chart 13. Do you consider that both machine learning and advertising processes automation will result in the reduced inclusion of
specialized personnel in advertising campaign management?
18
Source:
P
repared by the author
Fifty percent
of
interviewees did not hesitate to point out that companies will continue to rely on
human capital in an active and dynamic way to carry out certain parts or areas of the production
processes, compared to 30% who pointed out that the technology of algorithms will increasingly
have a greater weight even supplanting or replacing human presence. Finally, 20% said that this
reality would ultimately be shaped in accordance to the dynamics present in each company.
Chart 14. Is it dangerous or controversial to automate so many functions by including
advertising algorithms in platforms like Facebook and Google
?
Source:
P
repared by the author
Forty-
five percent of those interviewed indicated that there is no danger or risk involved in
algorithm presence. In contrast to this group, 35% of experts did not hesitate to affirm that for the
moment,
there is no clear or evident risk in task automation through advertising algorithms, and
finally, 20% indicated the notable presence of a risk or danger related to increased task
automation
.
Chart 15. Do you think that the trend of investing more and more in Facebook and
Google advertising will continue in the coming years or will companies return to
conventional advertising campaigns
?
ϱϬй
ϯϬй
ϮϬй
EŽ͘ ,ƵŵĂŶ ƌĞƐŽƵƌĐĞƐ ŝŶĐůƵƐŝŽŶ
ǁŝůů ĐŽŶƚŝŶƵĞ ƚŽ ďĞ ĂŶ ĞƐƐĞŶƚŝĂů
ĂƐƐĞƚ
zĞƐ͘ "ůŐŽƌŝƚŚŵŝĐ ƚĞĐŚŶŽůŽŐLJ ǁŝůů
ƵŶĚŽƵďƚĞĚůLJ ƉůĂLJ ĂŶ ŝŶĐƌĞĂƐŝŶŐůLJ
ŝŵƉŽƌƚĂŶƚ ƌŽůĞ͘
/ƚ ǁŝůů ĚĞƉĞŶĚ ŽŶ ĞĂĐŚ ĐŽŵƉĂŶLJ͕
ďƵƚ ŵĂĐŚŝŶĞƐ ǁŝůů ŶŽƚ ďĞ ĂďůĞ ƚŽ
ŽǀĞƌƚĂŬĞ ŚƵŵĂŶƐ͘
ϰϱй
ϮϬй
ϯϱй
EŽ͕ ďĞĐĂƵƐĞ ŝƚ ŝƐ ŶĞĐĞƐƐĂƌLJ ƚŽ
ƉĞƌĨŽƌŵ Ă ǁĞŝŐŚƚĞĚ ĂƐƐŝŐŶŵĞŶƚ
ŽĨ ƚĂƐŬƐ
zĞƐ͕ ďĞĐĂƵƐĞ ŵŽƌĞ ĂŶĚ ŵŽƌĞ
ĂĐƚŝŽŶƐ ĂƌĞ ƉĞƌĨŽƌŵĞĚ
ĂƵƚŽŵĂƚŝĐĂůůLJ
dŚĞƌĞ ŝƐ ŶŽ ĐůĞĂƌ Žƌ ĞǀŝĚĞŶƚ ƌŝƐŬ Ăƚ
ƚŚĞ ŵŽŵĞŶƚ
Source: Prepared by the author
Fifty percent of interviewees did not hesitate to point out that companies will continue to rely on human capital in an active
and dynamic way to carry out certain parts or areas of the production processes, compared to 30% who pointed out that the
technology of algorithms will increasingly have a greater weight even supplanting or replacing human presence. Finally, 20%
said that this reality would ultimately be shaped in accordance to the dynamics present in each company.
262
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Chart 14. Is it dangerous or controversial to automate so many functions by including advertising algorithms in platforms like
Facebook and Google?
18
Source:
P
repared by the author
Fifty percent
of
interviewees did not hesitate to point out that companies will continue to rely on
human capital in an active and dynamic way to carry out certain parts or areas of the production
processes, compared to 30% who pointed out that the technology of algorithms will increasingly
have a greater weight even supplanting or replacing human presence. Finally, 20% said that this
reality would ultimately be shaped in accordance to the dynamics present in each company.
Chart 14. Is it dangerous or controversial to automate so many functions by including
advertising algorithms in platforms like Facebook and Google
?
Source:
P
repared by the author
Forty-
five percent of those interviewed indicated that there is no danger or risk involved in
algorithm presence. In contrast to this group, 35% of experts did not hesitate to affirm that for the
moment,
there is no clear or evident risk in task automation through advertising algorithms, and
finally, 20% indicated the notable presence of a risk or danger related to increased task
automation
.
Chart 15. Do you think that the trend of investing more and more in Facebook and
Google advertising will continue in the coming years or will companies return to
conventional advertising campaigns
?
ϱϬй
ϯϬй
ϮϬй
EŽ͘ ,ƵŵĂŶ ƌĞƐŽƵƌĐĞƐ ŝŶĐůƵƐŝŽŶ
ǁŝůů ĐŽŶƚŝŶƵĞ ƚŽ ďĞ ĂŶ ĞƐƐĞŶƚŝĂů
ĂƐƐĞƚ
zĞƐ͘ "ůŐŽƌŝƚŚŵŝĐ ƚĞĐŚŶŽůŽŐLJ ǁŝůů
ƵŶĚŽƵďƚĞĚůLJ ƉůĂLJ ĂŶ ŝŶĐƌĞĂƐŝŶŐůLJ
ŝŵƉŽƌƚĂŶƚ ƌŽůĞ͘
/ƚ ǁŝůů ĚĞƉĞŶĚ ŽŶ ĞĂĐŚ ĐŽŵƉĂŶLJ͕
ďƵƚ ŵĂĐŚŝŶĞƐ ǁŝůů ŶŽƚ ďĞ ĂďůĞ ƚŽ
ŽǀĞƌƚĂŬĞ ŚƵŵĂŶƐ͘
ϰϱй
ϮϬй
ϯϱй
EŽ͕ ďĞĐĂƵƐĞ ŝƚ ŝƐ ŶĞĐĞƐƐĂƌLJ ƚŽ
ƉĞƌĨŽƌŵ Ă ǁĞŝŐŚƚĞĚ ĂƐƐŝŐŶŵĞŶƚ
ŽĨ ƚĂƐŬƐ
zĞƐ͕ ďĞĐĂƵƐĞ ŵŽƌĞ ĂŶĚ ŵŽƌĞ
ĂĐƚŝŽŶƐ ĂƌĞ ƉĞƌĨŽƌŵĞĚ
ĂƵƚŽŵĂƚŝĐĂůůLJ
dŚĞƌĞ ŝƐ ŶŽ ĐůĞĂƌ Žƌ ĞǀŝĚĞŶƚ ƌŝƐŬ Ăƚ
ƚŚĞ ŵŽŵĞŶƚ
Source: Prepared by the author
Forty-ve percent of those interviewed indicated that there is no danger or risk involved in algorithm presence. In contrast to
this group, 35% of experts did not hesitate to arm that for the moment, there is no clear or evident risk in task automation
through advertising algorithms, and nally, 20% indicated the notable presence of a risk or danger related to increased task
automation.
Chart 15. Do you think that the trend of investing more and more in Facebook and Google advertising will continue in the coming
years or will companies return to conventional advertising campaigns?
19
Source:
P
repared by the author
Fifty-
eight percent of respondents indicated that the dynamic of continued investment in Facebook
and Google advertising space will continue to grow and even increase. A total of 22% identified
unequivocally that investment in digital media depends mainly on each campaign.
Finally
, 20%
of respondents identified a scenario characterized by the presence of a hybrid model, i.e. a
situation halfway between online and conventional advertising.
Chart 16. Being a Facebook and Google partner, why do you think both platforms are
so opaque when it comes to how algorithms work
?
Source:
P
repared by the author
Fifty-
eight percent
of
interviewees considered that the lack of transparency in advertising
algorithms management is related to the fact that opacity is the key to these platforms’ success.
For 22% of those
interviewed
the main reason for the lack of clarity about how algorithms work
is that advertising management generates substantial profits for these companies. Finally, with
22%, a group of respondents who identified other reasons
.
Chart 17. Could Facebook and Google get more advertisers if they were more
transparent about advertising algorithms management
?
ϱϴй
ϮϬй
ϮϮй
zĞƐ͕ ƚŚŝƐ ǁŝůů ĐŽŶƚŝŶƵĞ ƚŽ ďĞ ƚŚĞ ĐĂƐĞ
ĂŶĚ ǁŝůů ĞǀĞŶ ŝŶĐƌĞĂƐĞ
EŽ͕ ŵŽƌĞ ĂŶĚ ŵŽƌĞ ĐŽŵƉĂŶŝĞƐ ĂƌĞ
ĐŚŽŽƐŝŶŐ ŚLJďƌŝĚ ĐĂŵƉĂŝŐŶƐ
/ƚ ǁŝůů ĚĞƉĞŶĚ ŽŶ ĞĂĐŚ ĐĂŵƉĂŝŐŶͲ
ďƵĚŐĞƚ ĂŶĚ ĂĚǀĞƌƚŝƐŝŶŐ
ŵĂŶĂŐĞŵĞŶƚ
ϱϴй
ϮϬй
ϮϮй
dŚĞ ůĂĐŬ ŽĨ ƚƌĂŶƐƉĂƌĞŶĐLJ ŚĂƐ ƚŽ ĚŽ
ŝŶ ƉĂƌƚ ǁŝƚŚ ĂĚǀĞƌƚŝƐŝŶŐ ƐƵĐĐĞƐƐ
!ĞĐĂƵƐĞ ŝĨ ƚŚĞ ĂůŐŽƌŝƚŚŵƐ ǁĞƌĞ
ĚŝƐĐůŽƐĞĚ ƚŚĞLJ ǁŽƵůĚ ĐĞĂƐĞ ƚŽ ĞdžŝƐƚ
dŚĞ ŵĂŝŶ ƌĞĂƐŽŶ ĨŽƌ ƚŚŝƐ ŝƐ ƚŚĂƚ
ĂĚǀĞƌƚŝƐŝŶŐ ŵĂŶĂŐĞŵĞŶƚ ŐĞŶĞƌĂƚĞƐ
ǀĞƌLJ ŚŝŐŚ ƉƌŽĨŝƚƐ
Source: Prepared by the author
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
263
Fifty-eight percent of respondents indicated that the dynamic of continued investment in Facebook and Google advertising
space will continue to grow and even increase. A total of 22% identied unequivocally that investment in digital media depends
mainly on each campaign. Finally, 20% of respondents identied a scenario characterized by the presence of a hybrid model,
i.e. a situation halfway between online and conventional advertising.
Chart 16. Being a Facebook and Google partner, why do you think both platforms are so opaque when it comes to how algorithms
work?
19
Source:
P
repared by the author
Fifty-
eight percent of respondents indicated that the dynamic of continued investment in Facebook
and Google advertising space will continue to grow and even increase. A total of 22% identified
unequivocally that investment in digital media depends mainly on each campaign.
Finally
, 20%
of respondents identified a scenario characterized by the presence of a hybrid model, i.e. a
situation halfway between online and conventional advertising.
Chart 16. Being a Facebook and Google partner, why do you think both platforms are
so opaque when it comes to how algorithms work
?
Source:
P
repared by the author
Fifty-
eight percent
of
interviewees considered that the lack of transparency in advertising
algorithms management is related to the fact that opacity is the key to these platforms’ success.
For 22% of those
interviewed
the main reason for the lack of clarity about how algorithms work
is that advertising management generates substantial profits for these companies. Finally, with
22%, a group of respondents who identified other reasons
.
Chart 17. Could Facebook and Google get more advertisers if they were more
transparent about advertising algorithms management
?
ϱϴй
ϮϬй
ϮϮй
zĞƐ͕ ƚŚŝƐ ǁŝůů ĐŽŶƚŝŶƵĞ ƚŽ ďĞ ƚŚĞ ĐĂƐĞ
ĂŶĚ ǁŝůů ĞǀĞŶ ŝŶĐƌĞĂƐĞ
EŽ͕ ŵŽƌĞ ĂŶĚ ŵŽƌĞ ĐŽŵƉĂŶŝĞƐ ĂƌĞ
ĐŚŽŽƐŝŶŐ ŚLJďƌŝĚ ĐĂŵƉĂŝŐŶƐ
/ƚ ǁŝůů ĚĞƉĞŶĚ ŽŶ ĞĂĐŚ ĐĂŵƉĂŝŐŶͲ
ďƵĚŐĞƚ ĂŶĚ ĂĚǀĞƌƚŝƐŝŶŐ
ŵĂŶĂŐĞŵĞŶƚ
ϱϴй
ϮϬй
ϮϮй
dŚĞ ůĂĐŬ ŽĨ ƚƌĂŶƐƉĂƌĞŶĐLJ ŚĂƐ ƚŽ ĚŽ
ŝŶ ƉĂƌƚ ǁŝƚŚ ĂĚǀĞƌƚŝƐŝŶŐ ƐƵĐĐĞƐƐ
!ĞĐĂƵƐĞ ŝĨ ƚŚĞ ĂůŐŽƌŝƚŚŵƐ ǁĞƌĞ
ĚŝƐĐůŽƐĞĚ ƚŚĞLJ ǁŽƵůĚ ĐĞĂƐĞ ƚŽ ĞdžŝƐƚ
dŚĞ ŵĂŝŶ ƌĞĂƐŽŶ ĨŽƌ ƚŚŝƐ ŝƐ ƚŚĂƚ
ĂĚǀĞƌƚŝƐŝŶŐ ŵĂŶĂŐĞŵĞŶƚ ŐĞŶĞƌĂƚĞƐ
ǀĞƌLJ ŚŝŐŚ ƉƌŽĨŝƚƐ
Source: Prepared by the author
Fifty-eight percent of interviewees considered that the lack of transparency in advertising algorithms management is related to
the fact that opacity is the key to these platforms’ success. For 22% of those interviewed the main reason for the lack of clarity
about how algorithms work is that advertising management generates substantial prots for these companies. Finally, with
22%, a group of respondents who identied other reasons.
264
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Chart 17. Could Facebook and Google get more advertisers if they were more transparent about advertising algorithms management?
20
Source: prepared by the author
For 50% of the interviewees, if Facebook and Google would disseminate more information about
how algorithms work and how useful they are, more and more companies would invest more in
online advertising. Some 30% of respondents said that both Facebook and Google would continue
to reach out to advertisers. For 20% of respondents, if Facebook and Google platforms were more
transparent, more and more companies would probably have a better understanding of how
algorithms work.
Chart 18. What characteristics should
o
nline and programmatic advertising experts
and campaign managers meet
?
Source:
P
repared by the author
For 35% of interviewees, algorithms and their knowledge require a series of specific skills that
are likely to be common among new professional profiles that may emerge. The same percentage
(35%) represented the experts and professionals who indicated with certainty the presence of new
profiles in medium and long term. Finally, for 30% of respondents, very competitive professional
profiles will appear.
Chart 19. Could conventional advertising agencies disappear due to online advertising
growth and its multiple advantages such as the ability to generate little budget
advertising?
ϱϬй
ϮϬй
ϯϬй
zĞƐ͕ ĞƐƉĞĐŝĂůůLJ ĨŽƌ ^D#Ɛ ĂŶĚ ĐŽŵƉĂŶŝĞƐ
ǁŝƚŚ ůŝƚƚůĞ ŬŶŽǁůĞĚŐĞ ŽĨ ŽŶůŝŶĞ ĂŶĚ
ƉƌŽŐƌĂŵŵĂƚŝĐ ĂĚǀĞƌƚŝƐŝŶŐ
WŽƐƐŝďůLJ ŝĨ ƚŚŝƐ ǁĞƌĞ ƚŚĞ ĐĂƐĞ ŵŽƌĞ ĂŶĚ
ŵŽƌĞ ĐŽŵƉĂŶŝĞƐ ǁŽƵůĚ ďĞƚƚĞƌ ƵŶĚĞƌƐƚĂŶĚ
ǁŚLJ ƚŽ ŝŶǀĞƐƚ ŝŶ ŽŶůŝŶĞ ĂĚǀĞƌƚŝƐŝŶŐ
EŽ͕ ďĞĐĂƵƐĞ &ĂĐĞŬ ĂŶĚ 'ŽŽŐůĞ
ĐŽŶƚŝŶƵĞ ƚŽ ƌĞĂĐŚ ĂĚǀĞƌƚŝƐĞƌƐ ĂŶĚ ǁŝůů ĚŽ ƐŽ
ƌĞŐĂƌĚůĞƐƐ ŽĨ ƚŚĞ ŝŶĨŽƌŵĂƚŝŽŶ ƚŚĂƚ ƚŚĞLJ
ƌĞůĞĂƐĞ ƚŽ ƚŚĞ ŽƵƚƐŝĚĞ
ϯϱй
ϯϱй
ϯϬй
"ůŐŽƌŝƚŚŵƐ ƌĞƋƵŝƌĞ ƐƉĞĐŝĨŝĐ ŬŶŽǁůĞĚŐĞ ƐŽ ŶĞǁ
ũŽď ƉƌŽĨŝůĞƐ ǁŝůů ĂƉƉĞĂƌ
EĞǁ ƉŽƐŝƚŝŽŶƐ ĂŶĚ ũŽď ŽƉƉŽƌƚƵŶŝƚŝĞƐ ǁŝůů
ĞŵĞƌŐĞ ĨŽƌ ƚŚĞ ŵŽƐƚ ĂĐƚŝǀĞ ƉƌŽĨŝůĞƐ
KŶůŝŶĞ ĂŶĚ ƉƌŽŐƌĂŵŵĂƚŝĐ ĂĚǀĞƌƚŝƐŝŶŐ ƌĞƋƵŝƌĞƐ
ĐŽŵƉĞƚŝƚŝǀĞ ƉƌŽĨŝůĞƐ ƚŚĂƚ ĂĚĂƉƚ ƚŽ ƚŚĞ ŶĞǁ
ĂĚǀĞƌƚŝƐŝŶŐ ĂŶĚ ƚĞĐŚŶŽůŽŐŝĐĂů ƌĞĂůŝƚŝĞƐ ĂƐ
ŽƉƉŽƐĞĚ ƚŽ ŵŽƌĞ ĐŽŶƐĞƌǀĂƚŝǀĞ ƉƌŽĨŝůĞƐ
Source: prepared by the author
For 50% of the interviewees, if Facebook and Google would disseminate more information about how algorithms work and how
useful they are, more and more companies would invest more in online advertising. Some 30% of respondents said that both
Facebook and Google would continue to reach out to advertisers. For 20% of respondents, if Facebook and Google platforms
were more transparent, more and more companies would probably have a better understanding of how algorithms work.
Chart 18. What characteristics should online and programmatic advertising experts and campaign managers meet?
20
Source: prepared by the author
For 50% of the interviewees, if Facebook and Google would disseminate more information about
how algorithms work and how useful they are, more and more companies would invest more in
online advertising. Some 30% of respondents said that both Facebook and Google would continue
to reach out to advertisers. For 20% of respondents, if Facebook and Google platforms were more
transparent, more and more companies would probably have a better understanding of how
algorithms work.
Chart 18. What characteristics should
o
nline and programmatic advertising experts
and campaign managers meet
?
Source:
P
repared by the author
For 35% of interviewees, algorithms and their knowledge require a series of specific skills that
are likely to be common among new professional profiles that may emerge. The same percentage
(35%) represented the experts and professionals who indicated with certainty the presence of new
profiles in medium and long term. Finally, for 30% of respondents, very competitive professional
profiles will appear.
Chart 19. Could conventional advertising agencies disappear due to online advertising
growth and its multiple advantages such as the ability to generate little budget
advertising?
ϱϬй
ϮϬй
ϯϬй
zĞƐ͕ ĞƐƉĞĐŝĂůůLJ ĨŽƌ ^D#Ɛ ĂŶĚ ĐŽŵƉĂŶŝĞƐ
ǁŝƚŚ ůŝƚƚůĞ ŬŶŽǁůĞĚŐĞ ŽĨ ŽŶůŝŶĞ ĂŶĚ
ƉƌŽŐƌĂŵŵĂƚŝĐ ĂĚǀĞƌƚŝƐŝŶŐ
WŽƐƐŝďůLJ ŝĨ ƚŚŝƐ ǁĞƌĞ ƚŚĞ ĐĂƐĞ ŵŽƌĞ ĂŶĚ
ŵŽƌĞ ĐŽŵƉĂŶŝĞƐ ǁŽƵůĚ ďĞƚƚĞƌ ƵŶĚĞƌƐƚĂŶĚ
ǁŚLJ ƚŽ ŝŶǀĞƐƚ ŝŶ ŽŶůŝŶĞ ĂĚǀĞƌƚŝƐŝŶŐ
EŽ͕ ďĞĐĂƵƐĞ &ĂĐĞŬ ĂŶĚ 'ŽŽŐůĞ
ĐŽŶƚŝŶƵĞ ƚŽ ƌĞĂĐŚ ĂĚǀĞƌƚŝƐĞƌƐ ĂŶĚ ǁŝůů ĚŽ ƐŽ
ƌĞŐĂƌĚůĞƐƐ ŽĨ ƚŚĞ ŝŶĨŽƌŵĂƚŝŽŶ ƚŚĂƚ ƚŚĞLJ
ƌĞůĞĂƐĞ ƚŽ ƚŚĞ ŽƵƚƐŝĚĞ
ϯϱй
ϯϱй
ϯϬй
"ůŐŽƌŝƚŚŵƐ ƌĞƋƵŝƌĞ ƐƉĞĐŝĨŝĐ ŬŶŽǁůĞĚŐĞ ƐŽ ŶĞǁ
ũŽď ƉƌŽĨŝůĞƐ ǁŝůů ĂƉƉĞĂƌ
EĞǁ ƉŽƐŝƚŝŽŶƐ ĂŶĚ ũŽď ŽƉƉŽƌƚƵŶŝƚŝĞƐ ǁŝůů
ĞŵĞƌŐĞ ĨŽƌ ƚŚĞ ŵŽƐƚ ĂĐƚŝǀĞ ƉƌŽĨŝůĞƐ
KŶůŝŶĞ ĂŶĚ ƉƌŽŐƌĂŵŵĂƚŝĐ ĂĚǀĞƌƚŝƐŝŶŐ ƌĞƋƵŝƌĞƐ
ĐŽŵƉĞƚŝƚŝǀĞ ƉƌŽĨŝůĞƐ ƚŚĂƚ ĂĚĂƉƚ ƚŽ ƚŚĞ ŶĞǁ
ĂĚǀĞƌƚŝƐŝŶŐ ĂŶĚ ƚĞĐŚŶŽůŽŐŝĐĂů ƌĞĂůŝƚŝĞƐ ĂƐ
ŽƉƉŽƐĞĚ ƚŽ ŵŽƌĞ ĐŽŶƐĞƌǀĂƚŝǀĞ ƉƌŽĨŝůĞƐ
Source: Prepared by the author
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
265
For 35% of interviewees, algorithms and their knowledge require a series of specic skills that are likely to be common
among new professional proles that may emerge. e same percentage (35%) represented the experts and professionals
who indicated with certainty the presence of new proles in medium and long term. Finally, for 30% of respondents, very
competitive professional proles will appear.
Chart 19. Could conventional advertising agencies disappear due to online advertising growth and its multiple advantages such as the
ability to generate little budget advertising?
21
Source:
P
repared by the author
For 40% of respondents, traditional advertising agencies will have to offer additional value or a
plus to avoid disappearance. Similarly, for 40% of respondents, the more conventional agencies
will have to distinguish themselves clearly from Facebook and Google platforms by offering
added value that will convince brands
Chart 20. What would be your recommendation, as an online advertising
professional
expert, for all profiles specializing in programmatic advertising and advertising
algorithms management
?
Source: prepared by the author
For 50% of interviewees, the importance of self
-
knowledge through a dynamic of self
-
training,
attendance at conferences or the like will be fundamental in terms of creating an expert profile.
This is followed by 30% of the interviewees who pointed out training relevance. Finally, 20% of
the respondents stressed the importance of continuous training in the face of a convulsive and
changing scenario.
5.
Discussion
and
Conclusions
After the methodological application, different conclusions have been obtained that need to be
analysed
and shared in order to verify whether the research objectives and hypotheses have been
met, or if, on the contrary, conclusive results have not been ach
ieved
.
ϰϬй
ϰϬй
ϮϬй
dƌĂĚŝƚŝŽŶĂů ĂĚǀĞƌƚŝƐŝŶŐ ĂŐĞŶĐŝĞƐ ǁŝůů
ŚĂǀĞ ƚŽ ŽĨĨĞƌ ĂĚĚŝƚŝŽŶĂů ǀĂůƵĞ ƚŽ ĂǀŽŝĚ
ĚŝƐĂƉƉĞĂƌĂŶĐĞ Žƌ ďĞĐŽŵŝŶŐ &ĂĐĞŬ
ĂŶĚ 'ŽŽŐůĞ ŶĞƚǁŽƌŬ ĂĚĚͲŽŶƐ
DŽƌĞ ĐŽŶǀĞŶƚŝŽŶĂů ĂŐĞŶĐŝĞƐ ŵƵƐƚ
ĐůĞĂƌůLJ ĚŝƐƚŝŶŐƵŝƐŚ ƚŚĞŵƐĞůǀĞƐ ĨƌŽŵ
ŽƚŚĞƌ ĂŐĞŶĐLJ ƚLJƉĞƐ ŐƌŽǁŝŶŐ ŝŶ
ƐƚƌĞŶŐƚŚ
dŚĞLJ ǁŝůů ŶŽƚ ĚŝƐĂƉƉĞĂƌ ďƵƚ ƚŚĞLJ ǁŝůů
ďĞ ƌĞĚƵĐĞĚ
ϯϬй
ϱϬй
ϮϬй
dŽ ďĞ ƚƌĂŝŶĞĚ ƚŚƌŽƵŐŚ
ŵĂƐƚĞƌΖƐ ĚĞŐƌĞĞ ƉƌŽŐƌĂŵƐ͕
ƐƉĞĐŝĨŝĐ ĐŽƵƌƐĞƐ ĂŶĚ ĐŽŶƚĞŶƚ
dŽ ďĞ ƐĞůĨ
Ͳ
ƚĂƵŐŚƚ ďLJ ƌĞĂĚŝŶŐ
ŵĂŶƵĂůƐ͕ ĂƚƚĞŶĚŝŶŐ
conferences…
dŽ ŬĞĞƉ Ă ĐŽŶƚŝŶƵŽƵƐ
ƚƌĂŝŶŝŶŐ ĂƚƚŝƚƵĚĞ ŝŶ Ă
ĐŚĂŶŐŝŶŐ ƐĐĞŶĂƌŝŽ͘
Source: Prepared by the author
For 40% of respondents, traditional advertising agencies will have to oer additional value or a plus to avoid disappearance.
Similarly, for 40% of respondents, the more conventional agencies will have to distinguish themselves clearly from Facebook
and Google platforms by oering added value that will convince brands
266
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Chart 20. What would be your recommendation, as an online advertising professional expert, for all proles specializing in
programmatic advertising and advertising algorithms management?
21
Source:
P
repared by the author
For 40% of respondents, traditional advertising agencies will have to offer additional value or a
plus to avoid disappearance. Similarly, for 40% of respondents, the more conventional agencies
will have to distinguish themselves clearly from Facebook and Google platforms by offering
added value that will convince brands
Chart 20. What would be your recommendation, as an online advertising
professional
expert, for all profiles specializing in programmatic advertising and advertising
algorithms management
?
Source: prepared by the author
For 50% of interviewees, the importance of self
-
knowledge through a dynamic of self
-
training,
attendance at conferences or the like will be fundamental in terms of creating an expert profile.
This is followed by 30% of the interviewees who pointed out training relevance. Finally, 20% of
the respondents stressed the importance of continuous training in the face of a convulsive and
changing scenario.
5.
Discussion
and
Conclusions
After the methodological application, different conclusions have been obtained that need to be
analysed
and shared in order to verify whether the research objectives and hypotheses have been
met, or if, on the contrary, conclusive results have not been ach
ieved
.
ϰϬй
ϰϬй
ϮϬй
dƌĂĚŝƚŝŽŶĂů ĂĚǀĞƌƚŝƐŝŶŐ ĂŐĞŶĐŝĞƐ ǁŝůů
ŚĂǀĞ ƚŽ ŽĨĨĞƌ ĂĚĚŝƚŝŽŶĂů ǀĂůƵĞ ƚŽ ĂǀŽŝĚ
ĚŝƐĂƉƉĞĂƌĂŶĐĞ Žƌ ďĞĐŽŵŝŶŐ &ĂĐĞŬ
ĂŶĚ 'ŽŽŐůĞ ŶĞƚǁŽƌŬ ĂĚĚͲŽŶƐ
DŽƌĞ ĐŽŶǀĞŶƚŝŽŶĂů ĂŐĞŶĐŝĞƐ ŵƵƐƚ
ĐůĞĂƌůLJ ĚŝƐƚŝŶŐƵŝƐŚ ƚŚĞŵƐĞůǀĞƐ ĨƌŽŵ
ŽƚŚĞƌ ĂŐĞŶĐLJ ƚLJƉĞƐ ŐƌŽǁŝŶŐ ŝŶ
ƐƚƌĞŶŐƚŚ
dŚĞLJ ǁŝůů ŶŽƚ ĚŝƐĂƉƉĞĂƌ ďƵƚ ƚŚĞLJ ǁŝůů
ďĞ ƌĞĚƵĐĞĚ
ϯϬй
ϱϬй
ϮϬй
dŽ ďĞ ƚƌĂŝŶĞĚ ƚŚƌŽƵŐŚ
ŵĂƐƚĞƌΖƐ ĚĞŐƌĞĞ ƉƌŽŐƌĂŵƐ͕
ƐƉĞĐŝĨŝĐ ĐŽƵƌƐĞƐ ĂŶĚ ĐŽŶƚĞŶƚ
dŽ ďĞ ƐĞůĨ
Ͳ
ƚĂƵŐŚƚ ďLJ ƌĞĂĚŝŶŐ
ŵĂŶƵĂůƐ͕ ĂƚƚĞŶĚŝŶŐ
conferences…
dŽ ŬĞĞƉ Ă ĐŽŶƚŝŶƵŽƵƐ
ƚƌĂŝŶŝŶŐ ĂƚƚŝƚƵĚĞ ŝŶ Ă
ĐŚĂŶŐŝŶŐ ƐĐĞŶĂƌŝŽ͘
Source: prepared by the author
For 50% of interviewees, the importance of self-knowledge through a dynamic of self-training, attendance at conferences or
the like will be fundamental in terms of creating an expert prole. is is followed by 30% of the interviewees who pointed out
training relevance. Finally, 20% of the respondents stressed the importance of continuous training in the face of a convulsive
and changing scenario.
5. Discussion and Conclusions
After the methodological application, dierent conclusions have been obtained that need to be analysed and shared in order
to verify whether the research objectives and hypotheses have been met, or if, on the contrary, conclusive results have not been
achieved.
–
Objective 1.-To analyse advertising algorithms management in Facebook ADS and Google AdWords as they are two of the
channels or supports that oer advertising most used by companies.
e rst research objective is to analyse the advertising algorithms operation of Facebook ADS and Google AdWords, two of
the most relevant platforms in online advertising today. is objective has been achieved not only with the methodological
section, but also with the bibliographical review of the study object. e survey and interviews with the research participants
have been fundamental to describe and analyse the advertising algorithms management. In parallel, research objective 1
could not be understood without the following starting hypothesis 1:
–
Hypothesis 1.-Facebook and Google algorithms perform better allocation of advertising campaign budgets compared to
manual allocation. e algorithms’ predictive capacity and impact are greater than human knowledge.
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
267
is starting hypothesis has been validated by taking into consideration the data obtained in the results phase after the
surveys and interviews application. For the professionals and experts consulted, the advertising algorithms used on Facebook
and Google platforms not only improve budget allocation in terms of campaigns, but also help to achieve better results by
impacting audiences.
–
Objective 2.- Describe the criteria used by algorithms, in addition to automated functionalities.
e second research objective is to describe the criteria used by algorithms and the automated functionalities present in
advertising algorithms. is research objective has not been 100% fullled due to the multiplicity, variety and diversity of
subjective data that advertising algorithms take into consideration to display a specic result for a specic search made by a
user.
We should not forget that algorithms oer personalized content based on the needs, search histories and demands made by
Internet users. erefore, records specialization is key. In relation to objective 2 of the work, hypothesis 2 has been described
in the following lines.
–
Hypothesis 2. Facebook and Google algorithms allow a higher level of analysis of ads created based on measurement,
impact and audience penetration criteria that are dicult to detect by professionals running advertising campaigns. is
hypothesis has been conrmed.
Reviewing the results shown in this work, it can be armed that advertising algorithms serve to eectively analyse the
behaviours that digital audiences generate in social networks, particularly Facebook, and in Google. us, it is important to
remember that campaigns’ segmentation and personalization can be achieved through manual action. However, the results
achieved do not reach the same level of personalization and user suitability.
–
Objective 3.-To address advertising algorithms dierences in terms of budget allocation, ad optimization and audience
detection.
e third and nal research objective aimed to address the divergences present in advertising algorithms in terms of budget
allocation and ad optimization. It can be considered that this objective has been met based on the results obtained and
the opinions expressed by the interviewees and respondents in considering that campaigns’ optimization is greater when
performing an automatic allocation than a manual one. However, budget allocation and audience detection depend too much
on the advertisers’ features, prole and typology.
Given this, and taking a specic example, it seems logical to understand that a sports retail sector multinational, with a
global presence, and a local SME specializing in multiservice provision, will have little or very little overlap in factors like
budget choice and new audiences prospecting. However, there is one element in common between the two companies: the
need to obtain better campaign results by optimizing ads, regardless of budget, duration or target audience. is leads to the
explanation of the starting hypotheses 3 and 4 set out in the following lines.
–
Hypothesis 3. Facebook and Google advertising algorithms serve to reach audience groups that are much related to pro
-
ducts, brands and services, if compared to the segmentation reached manually.
268
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
is hypothesis complements hypothesis number 2, so in both cases hypotheses have been fullled, arming that algorithms
are fundamental to achieve better campaign results, results optimization and emotional impact on audiences. is has been
conrmed by the data from the methodological application of the study.
–
Hypothesis 4. Facebook and Google’s advertising algorithms are one of the keys to these platforms’ success, which do not
clearly disclose the operation, information management, data and other elements compiled through the aforementioned
algorithms.
Hypothesis 4 has been rearmed in the research results phase. It can be particularly seen in the interview breakdown. In this
sense, the interviewees conrmed that management complexity, opacity and lack of transparency existing around Facebook
and Google’s advertising algorithms constitutes part of the success for both companies.
After describing the fullment of the research objectives and hypotheses, it is necessary to add that, according to the results
shown, the dierent respondents and interviewees consider that the future of advertising will include digital actions and
strategies such as programmatic advertising or online advertising as transcendental tools to ensure not only advertisers’
success in campaigns, but also to become a reference in today’s society. e various theoretical references consulted in the
research describe that online or programmatic advertising is consolidating as an increasingly useful and necessary model both
among advertising professionals and among consumers.
On the one hand, it is relevant to highlight that advertising algorithms inclusion by technological platforms like Facebook
and Google only draws a new scenario where academic and specic training is essential in terms of algorithm management.
is was underlined by the dierent respondents and interviewees who participated in the research shown and stressed the
importance of higher education and self-training for tomorrow’s advertising experts.
Another aspect of great relevance is that, despite the boom and growing importance of algorithms and programmatic advertising
in today’s society, these tools will never replace the manual and creative work done by human presence. However, it is not
wrong to consider that the inclusion of this type of technology in the realization, management and planning of advertising
helps to obtain better benets, both economic and of consumer impact, in terms of the advertising campaigns design.
On the other hand, an aspect of great relevance highlighted in this research and described both in the theoretical framework and
in the academic work results, aects the complexity and lack of transparency in Google and Facebook advertising algorithms.
In light of the above, the survey participants have not hesitated to state that a good part of these technological companies’
success derives from the advertising campaigns management through machine learning and articial intelligence inclusion.
is research work nally provides contrasted information on a study object in continuous change (online advertising and new
technologies), with multiple consequences, generating a new social, communicative and creative scenario characterized by
strategic decision-making. erefore, academic works like this one should be carried out in order to contribute to a clear vision
of what the advertising profession future will be like.
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
269
5. Acknowledgements
Mª Isabel Mansilla Blanco to whom the author is grateful for her work has translated this article into English.
6. Bibliographic References
Alonso, L.E. y Fernández, C.J. (2021). El papel del consumo en la economía de plataformas.
Revista Española de Sociología
(RES)
, 30 (3), a69. https://doi.org/10.22325/fes/res.2021.69
Alcalá, R. (2021). Evolución y análisis de la publicidad programática y estudio de la percepción del consumidor de la publicidad
online. Trabajo Final de Grado. Universidad de Valladolid. https://bit.ly/3x0EzEm
Barriola, J.M. et al. (2016). ¿Cómo funciona Google? El algoritmo Pagerank, diagrama de Grafos y cadenas de Markov.
Revista
de Investigación en Modelos Matemáticos Aplicados a la Gestión y la Economía
, número 3, 9-30. https://bit.ly/3z7WwUu
Boyd, J. (2018). Brandwatch. Objetivo de brandwatch. https://bit.ly/3wZ0CLW
Camargo-Vega, J.J., Camargo-Ortega, J. F. y Joyanes-Aguilar, L. (2014). Conociendo Big Data.
Revista Facultad de Ingeniería
,
enero-abril 2015, Vol. 24, número 38, 63-77. https://bit.ly/3aiWheH
Costalego, R. (2019). Aplicaciones de la inteligencia articial a la industria publicitaria. Trabajo Final de Grado. Universidad
de Valladolid. https://bit.ly/3PPEbBp
Chen, Y. et al. (2013). Social Media Metrics-A framework and guidelines for managing Social Media.
Journal of Interactive
Marketing
, Vol. 27, 281-289. https://doi.org/10.1016/j.intmar.2013.09.007
De Salas, M.I. (2010). La publicidad en las redes sociales. De lo invasivo a lo consentido.
Icono 14
, Vol. 8, 1. https://doi.
org/10.7195/ri14.v8i1.281
Del Rincón, D., Arnal, J., Latorre, A. y Sans, A. (1995).
Técnicas de Investigación en Ciencias Sociales
. Dykinson.
Denzin, N.K. y Lincoln, Y. S. (2005).
e sage handbook of qualitative research
. Sage.
Dermack, P. (2017). Aliate World Conferences. Entrevistado por Philipp Schoemann. Bangkok. https://bit.ly/3POKeGk
Díaz-Bravo, L., Torruco-García, U., Martínez-Hernández, M. y Varela-Ruiz, M. (2013). La entrevista, recurso exible y dinámico.
Investigación en Educación Médica
, Vol.2, núm. 7, 162-167. http://bitly.ws/e2gG
Enberg, J. (2019). Global Digital Ad Spending 2019. https://bit.ly/3t3AQoo
Estera, Á. (2016). 10 predicciones sobre el futuro del Big Data. https://bit.ly/3tki64b
Franks, B. (2012).
Taming the Big Data tidal wave: Finding opportunities in huge data streams with advanced analytics.
Wiley
and SAS Business.
Gaitán, J. A. y Piñuel J. L. (1998).
Técnicas de investigación en comunicación social. Elaboración y registro de datos.
Síntesis.
González, O. (2014).
Comercio electrónico 2.0
. Editorial Anaya.
Gupta, M. (2017). Machine Learning: preguntas y respuestas. https://bit.ly/3ax6Evr
270
|
nº 36, pp. 243-271
|
January-June of 2023
The management of advertising algorithms on the Internet. A case study: Facebook and Google
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Grané, J. (2021). ¿Jaque mate digital a la humanidad? Educar humanos y resilir en la era de la inteligencia articial.
DEDICA.
Revista de Educação e Humanidades
, núm. 18, 1-24. https://doi.org/10.30827/dreh.vi18.21000
Han, B.C. (2014).
Psicopolítica, Neoliberalismo y nuevas técnicas de poder
. Herder.
Iniesta-Alemán, I., Marta-Lazo, C. y Zaro, M. (2018). La inversión en publicidad, retos para la televisión del siglo XXI.
Retos:
Revista de Ciencias de la Administración y Economía
, Vol. 8, núm. 16, 141-156. https://doi.org/10.17163/ret.n16.2018.10
Jung, A. (2022).
Machine learning: the basics.
Springer.
Lucca, N. y Berrios, R. (2003).
Investigación cualitativa, fundamentos, diseños y estrategias
. Ediciones S.M.
Martínez, R. Segura R. y Sánchez, L. (2011). El complejo mundo de la interactividad: emociones y redes sociales.
Revista
Mediterránea de Comunicación
, Vol. 2, 171-190. https://doi.org/10.14198/MEDCOM2011.2.10
Martínez, I., Aguado, J.M. y Sánchez, P.H. (2022). Smart Advertising: innovación y disrupción tecnológica asociadas a la IA en el
ecosistema publicitario.
Revista Latina de Comunicación Social
, Vol.80, 69-90. https://www.doi.org/10.4185/RLCS-2022-1693
Mansky, M. (1968).
Semantic Information Processing
. e MIT Press.
Mitchell, T. M. (1997).
Machine Learning
. McGraw-Hill.
Montells, L. (2021). Qué es el Edgerank o cómo funciona el algoritmo de Facebook. https://bit.ly/38YPLJJ
Papi-Gálvez, N. et al. (2014).
Claves en la planicación de la publicidad online. Fundamentos, herramientas y retos
. AIMC
Asociación para la Investigación de Medios de Comunicación.
Peña, R. (2010). Historia de los algoritmos y de los lenguajes de programación. UNED. https://bit.ly/3t1irIP
Peña, R.M. (2017). Pensar la publicidad: una reexión losóca. Tesis doctoral. Universidad del País Vasco. San Sebastián.
https://bit.ly/3M1M9nZ
Pérez, R. M. (2020). El “dataísmo” como fundamento de la publicidad digital personalizada.
Ciencia y Sociedad
, número 45(4),
107-118. https://doi.org/10.22206/cys.2020.v45i4.pp107-118
Placebo Media (2016). RTB Placebo Media. https://bit.ly/3m265w9
Puyol, F. J. (2015).
Aproximación jurídica y económica al big data
. Tirant Editorial.
Ramos, M. (2006). Cuando la marca ofrece entretenimiento: aproximación al concepto de advertainment.
Questiones
Publicitarias
, I (1), 33-49. https://bit.ly/3M1dkPM
Raya, A. M. (2015).
Big Intelligence. Nuevas capacidades big data para los sistemas de vigilancia estratégica e inteligencia
competitiva
.
Rodríguez, R. (2021). El día que un ordenador derrotó a Kasparov: 25 años del “milagro” de Deep Blue. El Condencial. España.
https://bit.ly/38YmSxt
Rodriguez, A. (2017). La publicidad programática. El futuro de la publicidad digital. Trabajo Final de Grado. Universidad de
Extremadura. https://bit.ly/3lXg0mS
doxa.comunicación |
nº 36, pp. 243-271
|
January-June of 2023
Sergio Luque Ortiz
ISSN: 1696-019X / e-ISSN: 2386-3978
271
Saéz, C. (2020). El algoritmo como protagonista de la relación laboral. Un análisis desde la perspectiva de la prohibición de la
discriminación. Temas Laborales, núm.155-220, 41-60.
Sánchez, C. (2019). La publicidad digital superará a la televisiva en 2019. https://bit.ly/3PMBW1I
Sevillano, J. (2015). Comprando usuarios y no medios. Dos puntos sobre compra programática.
Anuncios: semanario de
publicidad y marketing
, número 1515, 8. https://bit.ly/3PMCapA
Steiner, C. (2012).
Automate this: How algorithms came to rule our world
. Editorial Portfolio.
Tobías, L., Vallejo, C. y Hinojo, P. (2022). El carácter disruptivo de la publicidad online y su impacto sobre la competencia.
ICE
Plataformas Digitales Regulación y Competencia
, 826, marzo-abril 2022.
Vallverdú, J.J. (2019). La publicidad digital superará a la televisiva en 2019. https://bit.ly/3N8bWwc
Vargas, I. (2012). La entrevista en la investigación cualitativa: nuevas tendencias y retos.
Revista Calidad en la Educación
Superior
, Vol.3, Núm. 1, pp.119-139. https://doi.org/10.22458/caes.v3i1.436
Villarreal, S. (2022). Publicidad programática: qué es y qué ventajas ofrece. IEBS https://www.iebschool.com/blog/
publicidad-programatica-marketing-digital/
Zuazo, N. (2018).
Los dueños de Internet.
Lectulandia.