Volume 6 (2020-21)

Each volume of Applied Marketing Analytics consists of FOUR 100-page issues, published in print and online.

The Articles published in Volume 6 include:

Volume 6 Number 4

Editorial: The process of winning with analytics: The emerging need for agility
Gordon Farquharson, Director, Metriculous
Practice papers
Data governance: The path to a data-driven culture
Stephanie Burton, Principal Consultant, Adobe Consulting

Companies have invested time, money and resources into expanding the depth and breadth of data available to them. However, most organisations struggle to use their data in their decision making. In order to enjoy a return on their investments, organisations need to create a data-driven culture. A data-driven culture is proven to improve customer acquisition, retention and profitability. The gap between having the data and using the data lies in data governance. This paper walks readers through four pillars of data governance. By breaking down each pillar into six key focus areas, readers can better understand how each pillar helps to create a data-driven culture.
Keywords: data governance, health, quality, collection strategy, digital maturity, organisational structure, democratisation, data-driven decisions
Transforming insight in 2021
James Wycherley, Chief Executive, Insight Management Academy

The COVID-19 pandemic has shone a spotlight on insight and analytics, with more executive interest in consumers and more focus on evidence-based decision-making. But does this mean that most organisations are insight-driven? The research suggests not; indeed, insight, analytics and research in most organisations remain fragmented and focused on tactical questions. The transformation of insight teams is the first step towards making organisations more insight-driven. To make this happen, this paper suggests a five-stage strategy: (1) identify value for the organisation; (2) drive change within the organisation; (3) lead insight strategy and people; (4) optimise impact through positioning and commerciality; and (5) accelerate the evolution of insight and then maintain momentum.
Keywords: insights, transformation, leadership, strategy, positioning, activation
Fixing the brand radar: Brand trackers should be used for management not market research
Niels Schillewaert, Managing Partner, InSites Consulting

Brand trackers have lost their credibility — they lack relevance, do not provide actionable insights, and can be problematic in terms of governance. The reason for this is simple: they are being used to the wrong ends. Brand trackers are tools for management, not for conducting research studies. They should function as a radar to detect competitive action in the marketplace so that executives can identify and solve issues before they become a problem. This paper proposes five measures to address this issue: (1) aligning the tracker with brand strategy; (2) measuring the content of topical creatives and events; (3) assessing the composition, effectiveness and efficiency of marketing mix touch points; (4) making the tracker adaptive and modular; and (5) mashing it up with the voice of the customer.
Keywords: brand tracking, branding research, consumer insights, advertising testing, voice of the customer, brand strategy
Capacity planning in marketing
Andrew Pearson, Managing Director, Intelligencia

Capacity management is the process of optimising production in line with fluctuating demand for products and services in order to reduce wasted capacity. Simply put, it aims to make optimal use of essential resources while minimising the use of non-essential resources. To this end, the key is to balance the right number of users and the right performance at peak usage to ensure a great end-user experience. This paper explores a capacity planning solution that can predict upcoming costs with advanced predictive analytics and forward-thinking what-if scenario modelling that can produce a healthy return on investment as well as help companies go green.
Keywords: capacity planning, real-time monitoring, AIOps, hyperautomation, application demand modelling, cloud bursting, predictive resource scaling, predictive analytics
The importance of domain knowledge for successful and robust predictive modelling
Andrea Ahlemeyer-Stubbe, Director Strategic Analytics and Agnes Müller, Senior Analytical Consultant, servicepro

Domain knowledge helps to build more precise and robust predictive models and thus obtain better insights. In the course of preparatory work, it helps inform what questions to ask, define the key fields to examine more closely, and identify where and how the insights from the analysis can support business goals. As this paper will discuss, it is also of great benefit when it comes to selecting or reducing variables, supplementing missing data, handling outliers or applying specific binning techniques. This paper argues that data scientists cannot rely on technical knowledge alone; rather, they must acquire relevant domain knowledge and familiarise themselves with pertinent rules of thumb. The paper also highlights the importance of maintaining close contact with the people who collect and prepare the data.
Keywords: predictive modelling, domain knowledge, binning, dummy variables, data preparation, missing data, data mining
Compiling user experience metrics via quantitative and qualitative methods
Patrick C. Leary, Vice President and User Experience Manager, AllianceBernstein

User experience as a discipline aims to apply intuitive approaches for users to engage with digital products. Defining foundational measurements and packaging recommendations clearly during syndication with stakeholders and peer teams allows strategy leaders to adjust as needed to maintain a holistic understanding of status and product direction while simultaneously making features more streamlined. This paper explores the benefits of assembling handoff material to measure and frame user experience measurements. It argues that there is a growing opportunity to improve upon digital product goals and ways to syndicate with teams across multiple disciplines.
Keywords: user experience, UX, key performance indicator, KPI, digital, marketing, product
In a data-rich world, which metrics should digital marketers focus on?
Simon Kingsnorth, Digital Marketing Consultant

The modern marketer has a wealth of data at their fingertips. However, being able to access vast quantities of data is not the same as knowing how to analyse and interpret said data. Indeed, the challenge is to review only the metrics that matter, or risk drowning in a sea of data. This paper discusses the landscape of modern marketing, to include the channels and core factors that every marketer must consider when deciding which key performance indicators to focus on. It also shows that by keeping an eye on their core strategy and understanding how to prioritise their data, every marketer can optimise their campaigns effectively.
Keywords: strategic dashboards, digital analytics, vanity metrics, web analytics, SEO data, social media analytics, ROI
Data protection and privacy in Canada: A balanced approach
Derek A. Lackey, Managing Director, Newport Thomson

This paper serves as an introduction on the upcoming Bill C-11 in Canada and discusses what marketers in Canada and those who want to do business there need to know about this bill.
Keywords: Bill C-11, data protection, data privacy, Canada, marketing, GDPR, California Consumer Privacy Act, PIPEDA
Research paper
Designing efficient assortments: A branch-and-bound method to optimise volume and satisfaction
Alessandro Martins Alves, Director, Ipsos, Marco Vriens, Chief Executive Officer, Kwantum and Thiago Graça Ramos, Project Manager, Ipsos

Category assortments are a key driver of retailer success. When a customer’s preferred product is unavailable, retailers must provide viable alternatives or risk losing the customer. This paper presents an analytical method that uses a survey-based approach where consumers make hypothetical (albeit not conjoint) choices. The objective is to identify how delisting a stock-keeping unit (SKU) will affect overall sales; the extent to which customers will abandon the store; and overall satisfaction with the store. Such knowledge will help retailers make the best delisting decisions and help both retailer and brands with their category management negotiations. It can also identify possible gaps for the brand, such as which alternatives in the competition’s portfolio have low substitutability and hence should perhaps be added to the brand’s product portfolio. This approach also allows brands to test new concepts and evaluate how adding a new (concept) SKU will affect sales of other SKUs and overall sales and profitability. The data are then modelled using a branch-and-bound algorithm. This approach can easily be implemented at the store level and allows for the optimisation of multiple objectives.
Keywords: assortment planning and optimisation, retail satisfaction, branch-and-bound optimisation

Volume 6 Number 3

Editorial: The process of winning with analytics: Provide clarity and confidence
Daniel Mooney, Editorial Board member
Practice papers
Using the customer journey to optimise the marketing technology stack
Seth Earley, CEO, Earley Information Science

This paper discusses how to evaluate marketing technology (martech) software in the context of the customer journey. It describes the technologies that are used to engage the customer at each stage of the customer journey, beginning with search engine optimisation to enable the inbound engagement that brings customers to a website and informs them about the available products or services. The paper describes the different characteristics of the customer journey, such as the fact that it traverses multiple departments and requires support from many organisational processes and structures. The paper concludes that in order to develop an appropriate martech software stack, it is essential to take a holistic view of the customer journey and the technologies used to support it. This paper provides a detailed analysis of an engagement strategy that spans the entire customer journey and identifies the technologies that are used for each stage, from search to purchase, order fulfilment, use of the product, and support. As implementing a full complement of solutions is not always possible, this paper provides guidance on how to prioritise the selection of enabling software, to guide the investment strategy by identifying those technologies that offer the greatest value to the organisation.
Keywords: CX, customer experience, martech, marketing technology, customer journey mapping, customer life cycle, metrics-driven governance, ROI, customer engagement technology evaluation
Pricing and proposition testing in subscription economies
Natasha Fosker, Senior Associate and Benny Cheung, Director, Decision Technology

As the subscription model continues to grow in popularity across diverse markets, this paper addresses three important issues in understanding the so-called subscription economy: (1) its prevalence; (2) the price sensitivity of different subscription types; and (3) whether promotional tactics can drive customer acquisition and retention. In addressing market prevalence, this paper reveals that 83 per cent of UK adults hold at least one type of subscription, with the most popular categories being service (71 per cent) and content (62 per cent), and the least popular being product subscriptions (30 per cent). Using the Behaviourlab platform, this paper finds all subscription industries are heavily price-driven; music and gym subscriptions are most price elastic of those tested, with price changes having double the impact on customer acquisition compared with other industries. The results from this study provide clear direction for future proposition design and the marketing of subscriptions.
Keywords: behavioural science, subscriptions, randomised controlled trials, RCT, consumer behaviour, acquisition, retention
Creating a ‘customer centricity graph’ from unstructured customer feedback
Elisabeth Lebmeier, Master’s Student and Naiwen Hou, Master’s Student, LMU Munich, Korbinian Spann, Managing Director, Insaas and Matthias Aßenmacher, PhD Student, LMU Munich

Certain industries, such as car insurance, do not have many customer touch points and do not offer a great deal of differentiation in the market. Marketers in such industries must therefore analyse vast amounts of customer-generated feedback in order to analyse customer preference in a quantitative manner. At present, this is done via market research or manual work, as an automated tool for summarising unstructured texts is as yet unavailable for certain European languages, including German. This paper discusses how Insaas and LMU Munich have used publicly available feedback on car insurance in Germany to develop a dedicated pipeline for the computation and visualisation of customer opinions. This paper provides an overview of the various steps of the procedure.
Keywords: customer centricity, NLP, AI, dashboard, B2C
Thesis and antithesis — Innovation and predictive analytics: Σ (Past + Present) Data ≠ Future Success
Ted William Gross, Founder, Asanatae

Predictive analytics (PA) is a tool routinely used by companies to help chart a future product path. It makes extensive use of algorithms and data mining to sort out market desires and trends. It also combines a robust host of artificial intelligence tools, including machine learning, pattern recognition, natural language processing, sentiment analysis and emotion recognition, among others, to achieve more precise results. PA, though, is imperfect, as it is often subject to the whims of the marketplace. Analysing past and present data does not, in any manner, guarantee positive results. Indeed, when it comes to innovation, particularly ‘disruptive innovation’, relying on PA can lead a company down a disastrous path. Data analytics requires a method that validates innovation and uses PA as something other than an infallible crystal ball. But does the possibility of innovation automatically disavow any insights into future market trends that PA may supply? This paper attempts to place both innovation and PA into proper perspective. It considers when, where, how and why PA and innovation are paramount, but reiterates the importance of instinct, originality and creativity. To illustrate its argument, the paper draws on the history of the Sony Walkman and Apple iPod.
Keywords: innovation, predictive analytics, disruptive innovation, disruption, market analytics, data analytics, artificial intelligence (AI)
Research papers
Design of experiments has not been used more widely by marketing despite its proven track record
Jonathan Frey, Co-Founder and Principal Consultant, Peninsula Business Intelligence and Harold S. Haller, Founder, Harold S. Haller & Company

In the global competitive marketplace, great products are coveted by consumers. Marketing continues to raise the bar on products that will delight customers and gain market share. Design of experiments is a well-established protocol for designing, developing and improving products and processes, which can be used to help development and production achieve marketing’s goals. The efficiency of experimental designs is a key feature that product developers and engineers can leverage in designing products in the resource-constrained business environment. The depth of information gained from following the experimental design protocol and the continual improvement cycle provides the developer or engineer great insights into what is important to the product and the customer. This paper presents three case studies to demonstrate the power of the experimental design protocol in developing products to meet marketing’s needs.
Keywords: experimental design, process improvement, optimisation
Why consumers fail to put their money where their mouth is: A study of organic coffee
Richard Friberg, Jacob Wallenberg Professor of Economics, Stockholm School of Economics and Mark Sanctuary, Senior Researcher, IVL Swedish Environmental Research Institute

Marketing research efforts often rely on surveys to determine what consumers want, complemented by data on observed purchases. The extent to which survey answers predict real market decisions, however, has been debated for decades. This paper investigates the relationship between survey responses and field data, using three years of household panel data on retail coffee purchases in Sweden. The data are obtained from an annual survey requiring households to indicate whether they try, to the extent feasible, to buy products that carry organic and Fairtrade labels. The results indicate that even households that say they try hard to purchase these products in fact buy mostly conventional coffee. This paper investigates this discrepancy using an estimated structural demand model of household coffee purchases that combines both survey and field data. The study finds that hardcore organic (Fairtrade) households want to buy organic (Fairtrade) coffee but do not because it is not available from their preferred brand. Moreover, the high price of organic coffee is the main deterrent for households with a more moderate desire to purchase organic/Fairtrade.
Keywords: combining stated and revealed preferences, consumer scan panel, eco-labelling
The benefits of Shapley-value in key-driver analysis
Marco Vriens, Chief Executive Officer, Kwantum, Chad Vidden, Associate Professor of Mathematics and Statistics, University of Wisconsin-La Crosse and Nathan Bosch, Master’s Student, KTH Royal Institute of Technology

Linear (and other types of) regression are often used in what is referred to as ‘driver modelling’ in customer satisfaction studies. The goal of such research is often to determine the relative importance of various sub-components of the product or service in terms of predicting and explaining overall satisfaction. Driver modelling can also be used to determine the drivers of value, likelihood to recommend, etc. A common problem is that the independent variables are correlated, making it difficult to get a good estimate of the importance of the ‘drivers’. This problem is well known under conditions of severe multicollinearity, and alternatives like the Shapley-value approach have been proposed to mitigate this issue. This paper shows that Shapley-value may even have benefits in conditions of mild collinearity. The study compares linear regression, random forests and gradient boosting with the Shapley-value approach to regression and shows that the results are more consistent with bivariate correlations. However, Shapley-value regression does result in a small decrease in k-fold validation results.
Keywords: driver modelling, regression, Shapley-value, customer satisfaction, random forests, gradient boosting
Online hotel reviews: Factors influencing how customers perceive their credibility and the likelihood of customer adoption
Talal Abuhulaibah, PhD Student, Oklahoma State University, Shiang-Lih Chen McCain, Assistant Professor of Marketing, Colorado Mesa University and Jeffrey C. Lolli, Associate Professor, Widener University

Customer reviews are an increasingly important part of the travel and tourism industry. It is therefore important to understand how the perceived credibility of online reviews influence the likelihood of customer adoption. This study examines three factors that affect the credibility of online reviews and how those reviews influence customer adoption when booking hotel reservations, namely: argument quality, source credibility, and review consistency and quantity. The results show that these factors have direct and positive effects for both hotel managers and online booking and review websites. Hotel managers should assess all reviews for credibility by looking at the argument and source credibility, and then look for trends or patterns to uncover recurring issues. Most importantly, managers must respond to all reviews in a prompt, helpful and polite manner. This can influence not only the individual who has posted the online review, but also the many readers thereafter. Furthermore, online booking and review sites can do more to ensure fair postings and increase the credibility of online hotel reviews. These sites should not allow anonymous postings but rather require each reviewer to create a profile that includes ID verification.
Keywords: online hotel reviews, online review credibility, argument quality, source credibility, review consistency and quantity, online review adoption

Volume 6 Number 2

Editorial: Analytics with a conscience: Values-based decision making within marketing analytics
Ted Kwartler, Professor, Harvard University Extension School
Practice papers
Hiring strategies beyond the pandemic: Attracting the best marketing analytics and business intelligence talent
Kate D’Alessandro, Manager of Marketing and Operations and Steve Perlman, Co-founder and CEO, Syfter

As technology advances, the need for digitally adept professionals within the marketing sphere increases, especially within the fields of analytics and business intelligence. This paper provides a closer look at the most effective ways to understand and fill the needs of marketing teams, and highlights the need for creative recruitment strategies.
Keywords: staffing, recruiting, marketing, analytics, business intelligence, strategies
Response latency measures in questionnaires: A brief overview
Darren Bridger, Founder, CloudArmy Neuro

Questionnaire response times have been used in psychological research for decades. However, even though the prevalence of computers has made such data more easily available than ever, market research frequently overlooks this information. Adding response timing to survey questions can increase both their reliability and their ability to predict behaviour. They can also be indicative of attitudes that are more resistant to change. Faster response times are thought to indicate attitudes that are highly accessible from memory, but they are only measurable under certain conditions. Non-attitudinal variables that can also affect response speed must also be considered. This paper reviews the evidence supporting the advantages of adding response time data to surveys, and briefly introduces an example procedure for doing so.
Keywords: response-latency, attitude accessibility, response-time, System 1, dual-choice
A detective’s story: How tiny clues can reveal great opportunities
Iulia Cornigeanu, UX Research Manager, On the Beach Ltd

Big Data and advanced analytics are powerful tools for any data-driven team. Nevertheless, there are times when relying on vast amounts of data or dry A/B testing will deliver only a partial image. In such cases, the quality of this picture could be improved by embracing tactics such as avoiding wilful blindness, challenging preconceptions and adopting hybrid, bespoke approaches that deliver both the what and the why. This paper describes how ‘small data’ (to use Lindstrom’s term) can save the day by helping translate numbers into action, revealing a massive opportunity that would otherwise be missed.
Keywords: data-driven, customer-centric, insights, small data, split testing (A/B testing), qualitative research, wilful blindness
Customer relationship management with WeChat: How to apply analytics principles to social media in China
Arnold Ma, CEO/Founder, Qumin

WeChat is by far the most important social communication platform in China, encompassing more than a billion daily active users across an entire ecosystem of touch points. It combines peer-to-peer messaging, online payments, business accounts, miniprograms, social sharing and business productivity. This presents great opportunities for brands looking to do business in China as well as serious challenges. This paper contextualises WeChat’s development from app to superapp since its launch in 2011. It looks at how the WeChat account itself is changing and what that this indicates about the average Chinese netizen. The paper then delineates the changes in WeChat and how marketing and sales people are leveraging them, including some of the most common tactics for conducting successful customer relationship management via WeChat. Finally, the paper discusses how WeChat is likely to develop in the coming years.
Keywords: CRM, China, WeChat, analytics, social media
Research papers
Will they stay or will they go? Predicting customer churn in the energy sector
Michela Vezzoli, Postdoctoral Research Fellow, Cristina Zogmaister, Associate Professor in Psychometrics, University of Milano Bicocca and Dirk Van den Poel, Full Professor of Marketing, Ghent University

The liberalisation of the European energy market has driven changes in the way firms approach marketing, both for the acquisition of new consumers and for retaining existing ones. To retain consumers, practitioners aim to predict which consumers intend to churn (ie leave), and to understand the reasons behind this intention. To address this need, this study uses data-mining techniques to develop a churn prediction model. The study aims to identify the information that is predictive of churn and, consequently, to shed light on the psychological reasons behind churn. The authors built eight predictive models using decision trees, random forest and logistic regression on a dataset composed of 81,813 consumers of an energy provider, each with one residential electricity contract. The logistic regression was found to outperform the other methods. The discussion focuses on the relevant predictors of churn by addressing a posteriori psychological explanations of consumers’ churn behaviour. The study provides new insights on the reasons why customers churn and, by addressing theoretical psychological explanations, provides a data-mining model with robustness to contextual changes.
Keywords: churn prediction model, customer churn, consumer psychology, machine learning, energy market
An insight into customers’ product return intentions
Devdeep Maity, Associate Professor, Delaware State University

Every year, the global retail industry loses billions of dollars due to product returns. Despite this, very little of the marketing literature examines customers’ intentions to return and the associated implications. The present study captures, analyses and interprets a managerially relevant variable, namely product return intentions, and helps retail store managers, retail analysts and marketing managers understand the decision journey behind consumer product returns. Specifically, the study looks at the relationship between the consumer’s cognitive dissonance after purchase and product return intentions in the context of both lenient and strict return policies. The study also looks at the ill effects of such return intentions in terms of negative word-of-mouth feedback about the retailer. The results suggest that cognitive dissonance after purchase positively influences product return intentions, which in turn positively influences negative word-of-mouth intentions; however, the relationship remains unaffected in strict and lenient return policy situations. Besides providing crucial insights into the complex consumer sentiments associated with seemingly simple product returns, the research questions the previously held notion of retailers and store managers controlling product returns by imposing strict return policies — a finding of paramount importance to business.
Keywords: product return intentions, cognitive dissonance after purchase, return policy leniency, negative word-of-mouth intentions
The influence of social media communities on brand loyalty: Case study of Morocco’s telephone operating companies
Kerkri Abdelmounaim, Professor, School of Higher Studies in Engineering, Morocco

The telecommunication industry in Morocco has witnessed substantial growth in the past decade, with three main brands dominating the market. This paper assesses the efficiency of these brands’ social media presence by modelling the relationship between social media, brand loyalty and brand trust. The study uses partial least squares path modelling to test six hypotheses. The results suggest that social media play a significant role in building brands in the Moroccan telecommunication sector.
Keywords: social media, brand loyalty, brand trust, PLS path modelling, telecommunication

Volume 6 Number 1

Editorial: Marketing science and epidemiology
Dominique M. Hanssens, Editorial Board Member, Applied Marketing Analytics
Using event-based models for cross-device insights into the user journey
Mai Alowaish, Digital Marketing Consultant and Delivery Team Lead, InfoTrust

Gaining customer insight is the foundation of marketing science. Today’s customers interact with companies across multiple devices. Thus, to fully understand the customer’s journey, digital marketing analysts can no longer rely on measuring individual events on a single device; rather, they must gather and interpret data across devices. This requires making the user the unit of data measurement. The challenge is stitching together information into a single story and profile that lead to actionable and immediate insights. This paper discusses event-based models for digital marketing data and calls for marketing analytics leaders to execute the transformation from static session-based reporting to dynamic and powerful user-centric analytics.
Keywords: digital analytics, web analytics, app analytics, event-data models, Internet-of-Things, cross-device
Case study: In-store display and visual merchandising analytics
Gary Angel, Founder and Chief Executive Officer, Digital Mortar

In this project, video and advanced machine learning were used to analyse shopper behaviour in a key display area at multiple locations of a multi-billion-dollar retailer. The goal was to understand the volume of shoppers using the area, engagement with the displays, and whether the displays generated product interactions and takeaways. In addition, the system was used to design and support an aggressive testing programme to optimise the display area. Measurement answered the usage questions and revealed obvious opportunities for improvement. Structured testing revealed that the geometry of the area heavily impacted usage and engagement, that shopper flow was strongly influenced by changing the density and alignment of the features, and that there were opportunities for improving product mix and layout. As the purpose of a store is to get shoppers’ eyes and hands directly on product, the ability of product displays to attract and engage shoppers is critical to retail success. Like many aspects of physical retail, however, merchants have little visibility into the success of any given display and insufficient measurement to drive testing and improvement programmes. This case study shows how measurement and testing in display has become possible using people-counting technologies.
Keywords: retail analytics, store analytics, merchandising, display, endcap, display measurement, display performance, display engagement
Sentiment analysis and emotion recognition: Evolving the paradigm of communication within data classification
Ted William Gross, AI Technologist and Data Theorist, Ituran

The process of sentiment analysis and emotion recognition (SAER) entails using artificial intelligence components and algorithms to extract emotions and sentiments from online texts, such as tweets. The information extracted can then be used by marketing, customer support and public relations teams to foster positive consumer attitudes. Advances in this discipline, however, are being hindered by two significant obstacles. First, although ‘emotion’ and ‘sentiment’ are distinct entities that require distinct analysis, there is no agreed definition to distinguish between the two. Secondly, the nature of language within the electronic medium has evolved to include much more than textual statements, including (but not limited to) acronyms, emojis and other visuals, such as video (in its many forms). As visual communication lacks universal interpretation, this can lead to erroneous analysis and conclusions, even where there is a differentiation between emotion and sentiment. This paper uses examples and case studies to explain the theoretical basis of the problem. It also offers conceptual direction regarding how to make SAER more accurate.
Keywords: sentiment analysis, emotion recognition, contextual analysis, communication, emoji symbolisation, data analytics
Facebook and Pandora’s box: How using Big Data and Artificial Intelligence in advertising resulted in housing discrimination
Sarah Khatry, Data Scientist and Writer, DataRobot

In 2019, the US Department of Housing and Development charged Facebook with violating the Fair Housing Act 1968. This followed an investigation into the use of ethnically targeted advertising practices on Facebook. To understand Facebook’s targeting methods and the cause of the problematic outcomes, this paper follows the journey of an advertisement through Facebook’s platform. In this way, Facebook’s regulatory missteps can serve as a case study to illustrate how Big Data analytics can, when informed by human and machine bias, cross the line into discriminatory practices. This case study underscores how it is vital — in advertising as in other industries — not to treat advanced analytics like artificial intelligence as black boxes. Indeed, to inform the design and use of advanced analytics, it is essential for companies to consistently develop a comprehensive understanding of their data, in addition to the legal and ethical implications of their operations.
Keywords: housing discrimination, Facebook, Big Data, AI, targeted advertising
Building analytics teams for success
Radhika V. Kulkarni, Former Vice President of Advanced Analytics R&D, SAS Institute

Big Data, analytics, Artificial Intelligence, data science: all of these areas have given rise to the need for every industry to invest in building up analytics teams that can address complex business challenges. This paper looks at some aspects that are critical for the composition of successful teams in terms of technical background, skills and experience. It also discusses the challenges companies face in hiring the right individuals for these varied roles and the role educational institutions play in filling this talent gap.
Keywords: analytics excellence, successful analytical groups, analytical skills gap, democratisation of analytics, marketing optimisation, data science, Artificial Intelligence
Four pivotal capabilities for marketers to support the growth mandate
Laura Patterson, President, VisionEdge Marketing

Whether times are uncertain or our sails are full of wind, there is always the pressure to grow. Every company must take a customer-centric approach to stay on track or to ramp up. This paper explores four pivotal capabilities that every company needs their marketing organisation to master in support of the growth mandate.
Keywords: customer-centric, organisation, market growth, data-driven, analytics, marketing, strategy
Adapting the enterprise data lake architecture for marketing analytics
Roger Kamena, Vice-President of Data Science and Technology, Adviso Conseil

Data lakes have evolved over the last decade from highly complex IT infrastructures on-premise to simpler serverless cloud environments. This change makes data lakes much less difficult to maintain and operate, which provides an opportunity to democratise them beyond operational business intelligence teams to a wider audience, namely marketing analytics practitioners. At the same time, the availability of marketing data has also evolved over the last decade. Marketing technology platforms have proliferated the variety, velocity and volume of data for marketers to process. This paper considers how a data lake architecture could be adapted for marketers to help them go further with their data.
Keywords: data lake, data management systems, Big Data analytics, cloud analytics, marketing data, data architecture
Interrelated factors driving the purchase of over-the-top television subscription services: A study using exploratory factor analysis and the decision-making trial and evaluation laboratory method
Chand P. Saini, Assistant Professor, SGT University and Neha Gupta, Assistant Professor, Amity School of Business

This study identifies the key factors motivating people to subscribe to over-the-top (OTT) services and examines the causal relationships between those factors. As it is a multiple criteria decision-making problem, the study embraces an integrated model of exploratory factor analysis and the decision-making trial and evaluation laboratory (DEMATEL) method. The latter technique is used to simplify the decision-making process and visualise the interrelationships between the motivation factors. The study identifies four key motivational factors: flexibility, content, variety and social status, of which content is the most crucial. The findings of this study will be of interest to marketing analysts and decision-makers seeking to measure, analyse and improve marketing performance in order to develop better strategies for marketing OTT services.
Keywords: decision-making, OTT, live-streaming, flexibility, content, factor analysis, DEMATEL
Book review: The AI-Powered Enterprise: Harness the Power of Ontologies to Make Your Business Smarter, Faster and More Profitable
Reviewed by Denis Malin, Editorial Board Member
Book review: Fast Track Your Business: A Customer-Centric Approach to Accelerate Market Growth
Reviewed by Seth Earley, Editorial Board Member