CHAPTER 4 | Managing Marketing Information to Gain Customer Insights 133

Behavioral and Social Targeting: Sophisticated

Real Marketing 4.1 Marketing or Just a Little Creepy?
Thanks to the burgeoning world of web brows- running shoes (behavioral targeting), an ad to purchase ads that target the right cus-
ing, social media, mobile apps, online shopping, for a specific pair of running shoes pops up tomers. So if you’re browsing lawn and gar-
and other internet activities, marketers now have because a friend that you’re connected to via den sites, don’t be surprised to see ads for
real-time access to a flood of online consumer Instagram or Twitter just bought those shoes Scotts lawn products the next time you visit
information. It’s all there for the digging—what from Zappos.com last week (social targeting). Weather.com. Or if you seek car-buying ad-
sites consumers visit, what searches they make, Social targeting can even capture the dy- vice at sites such as Edmunds.com or nada-
what apps they use, how they shop, what they namics of real-time conversations. For exam- guides.com, you might well see ads for the
buy, with whom they interact—digitally revealed ple, beyond just targeting 24- to 26-year-old cars you researched the next time you visit
as they navigate the internet. males who are both sports fans and car enthu- Google News to catch up on what’s happen-
Marketers routinely employ sophisticated siasts, Chevrolet made its ad message even ing around the nation and world.
big data tools to analyze the churning mass of more relevant by targeting those consumers The major social media have also got-
online and mobile data in precise detail, using while they are talking about football on a mo- ten into behavioral targeting in a big way.
the resulting insights to target and personalize bile Twitter app during the Super Bowl. When Facebook, Google, Instagram, Twitter,
marketing ads and offers. On today’s internet, they checked the app, targeted consumers Snapchat, and other social media dig deeply
everyone knows who you are. By combining saw an ad that prompted them to check out into user data to help their advertisers target
online and offline data, marketers know your Chevy’s Super Bowl video on YouTube. more sharply. For example, some 2 billion
age, your gender, where you live, that you love Behavioral and social targeting require so- people use Facebook every month. However,
dogs, what you bought recently at Amazon. phisticated analytics, so many marketers enlist based on sophisticated analytics, Facebook
com, and that you spent one hour and 21 min- the services of specialized ad networks with offers powerful audience selection tools that
utes last Sunday morning browsing college exotic names like Taboola, PulsePoint, and help its advertisers target the right customer
basketball news and scores at ESPN.com. Adknowledge. These digital advertising net- groups or even individuals on Facebook.
Marketers use all that data to deliver ads works obtain user browsing data by partner- Advertisers can target Facebook users
and offers aimed squarely at individual con- ing with hundreds or even thousands of web- based on demographics (age, gender, edu-
sumers, wherever they travel on the inter- sites. The partners supply gobs of data on user cation, relationship status, or even job titles),
net, or even in stores. It’s called behavioral browsing histories, web and mobile site usage, location (where they live or maybe within a
targeting—tracking consumers’ online behav- electronic shopping cart contents, and other radius around a store), interests (say, hobbies
ior and using it to target ads and offers to them. details of what users do, where, and when. or favorite entertainment), or behaviors (what
So, for example, if you do a Google search for The ad networks then apply high-powered they buy, device usage, or other activities). Or
a Samsung TV you’re thinking about buying, big data analytics to identify consumers with Facebook can help advertisers create “cus-
you’ll probably see an ad for that very type similar interests, needs, behavior, and inter- tom audiences” by finding and reaching ex-
of TV on your next visit to Facebook or your net habits. This lets them merge audience isting customers and contacts who also use
favorite buying site. Or as you’re shopping data from one group of sites with ad place- Facebook. Advertisers can even build what
in one section of your local Walgreen’s, you ments on another. Armed with that informa- Facebook calls “lookalike audiences,” people
might receive a real-time notification on your tion, the networks then work with advertisers on Facebook whose behaviors mirror those
phone of a deal in another
section of the store.
All this is amazing enough, SEARC
HE INE:
but web analytics and target- In-hous D ONLINE F ED ONL deos
e perso O WATCH orkout vi
ing take online eavesdropping nal train R: e a n d w
ers Exercis
even further—from behavioral
targeting to social targeting.
Whereas behavioral target-
ing tracks consumer move- CLICKED ON A LINK TO: SHARED ON SOCIAL MEDIA:
ments across online sites, Outdoor gear and apparel Hiking-with-friends photos
social targeting also mines
individual online social media
connections and conversa- REA
R: D
ES FO rs Natu BLOG
tions. Research shows that IT ral a P
D W E B S
b r id c a nd o OSTS O
E y rgan
consumers shop a lot like their VISIT ompact h ic foo N:
C ds
friends and are five times more
likely to respond to ads from
Other data such as demographics, psychographics, geo-location,
brands friends use. Social tar-
and previous purchases useful for behavioral and social targeting.
geting links customer data to
social interaction data from GENDER: RECENT PURCHASES: LOCATION:
social networking sites. Female Fitness tracker, MacBook Pro California
So, instead of just having
a Zappos.com ad for run- On today’s internet, with today’s advanced analytics, everyone knows who you are. Marketers use such
ning shoes pop up because insights to personalize online shopping experiences. But is it sophisticated marketing or “just a little creepy”?
you’ve recently searched for Rido/Shutterstock
134 PART 2 | Understanding the Marketplace and Consumer Value
of their target customers. Thanks to such so- Yup, that’s the downside. At what point does Despite such concerns, behavior and
phisticated targeting capabilities, Facebook sophisticated online research cross the line social targeting continue to grow and to get
now captures more that 15 percent of all into unwelcome invasions of privacy? smarter. And, with appropriate safeguards,
online advertising dollars, second only to Proponents claim that behavioral and social they promise benefits for both companies
Google’s more than 40 percent. targeting benefit more than abuse consumers and customers. Marketers who can plumb
Online analytics. Behavioral targeting. by feeding back ads and products that are the depths of customer data and respond in
Social targeting. All of these are great for mar- more relevant to their interests. But to many meaningful, valuable ways without crossing
keters as they work to mine customer insights consumers and public advocates, following the line will thrive. But it can be a delicate bal-
from the massive amounts of consumer in- consumers online and stalking them with ads ance. As Snapchat’s Privacy Center states,
formation swirling around the internet. The feels more than just a little creepy. “Marketers “We want you to feel understood. We want to
biggest question? You’ve probably already must find a sweet spot between personal- understand what’s relevant to you and your
guessed it. As marketers get more adept at ization and surveillance,” says one analyst. life, and we want to show you things that
trolling web and mobile sites, social media, “Customers can be creeped out when brands you’ll care about. At the same time, we don’t
and other digital domains, what happens to cross the thin line between knowing them like want to serve ads that are so custom-tailored
consumer privacy and personal data security? a friend and knowing them like a stalker.” that they feel invasive or uncomfortable.”24

are many ways to ask questions. Closed-ended questions include all the possible answers,
and subjects make choices among them. Examples include multiple-choice questions and
scale questions. Open-ended questions allow respondents to answer in their own words.
In a survey of airline users, Southwest Airlines might simply ask, “What is your opinion
of Southwest Airlines?” Or it might ask people to complete a sentence: “When I choose
an airline, the most important consideration is...” These and other kinds of open-ended
questions often reveal more than closed-ended questions because they do not limit respon-
dents’ answers.
Open-ended questions are especially useful in exploratory research, when the re-
searcher is trying to find out what people think but is not measuring how many people think
in a certain way. Closed-ended questions, on the other hand, provide answers that are
easier to interpret and tabulate.
Researchers should also use care in the wording and ordering of questions. They should
use simple, direct, and unbiased wording. Questions should be arranged in a logical order.
The first question should create interest if possible, and difficult or personal questions
should be asked last so that respondents do not become defensive.

Mechanical Instruments. Although questionnaires are the most common research in-
strument, researchers also use mechanical instruments to monitor consumer behavior.

Table 4.2 | Types of Samples

Probability Sample

Simple random sample Every member of the population has a known and equal chance of selection.

Stratified random sample The population is divided into mutually exclusive groups (such as age groups), and random
samples are drawn from each group.

Cluster (area) sample The population is divided into mutually exclusive groups (such as blocks), and the researcher
draws a sample of the groups to interview.

Nonprobability Sample

Convenience sample The researcher selects the easiest population members from which to obtain information.

Judgment sample The researcher uses his or her judgment to select population members who are good prospects
for accurate information.

Quota sample The researcher finds and interviews a prescribed number of people in each of several categories.
 CHAPTER 4 | Managing Marketing Information to Gain Customer Insights 139

Artificial Intelligence in Marketing: “A Bigger

Real Marketing 4.2 Deal Than Fire or Electricity”
It’s early morning, you’re headed out to start Marketers use AI to assess, address, ser- such as Google Home, have become voice-
your day, and you feel the urge for that first vice, and sell to customers. In turn, AI can activated personal shopping assistants.
jolt of caffeine. As you get in your car, you help customers manage their lives and their Companies ranging from P&G and Clorox to
tap the Starbucks app on your phone and buying. It might be requesting a ride from Lyft 1-800-Flowers are hard at work perfecting
ask for “the usual.” Your Starbucks virtual via chat (Facebook Messenger or Slack) or ways to tap into Echo users who voice-shop
barista replies in her familiar, cheerful voice: voice (Amazon Echo’s Alexa virtual assis- from the comfort of their own kitchens.
“One tall caramel latte!” She then politely sug- tant). Lyft’s chatbot lets you know the current At Amazon’s shopping and video sites,
gests a breakfast snack—a Vermont maple location of your driver along with a picture of AI powers recommendations that help con-
nut muffin—not your usual, but it sure sounds the license plate and car model. Or it might sumers decide what to buy and what to
good. You agree. “Thanks! Your order will be be IBM’s Watson supercomputer combing watch. “Increasingly, Amazon will be sell-
ready for pickup in five to seven minutes at through vast amounts of data to unearth ing you things you didn’t even know you
University and 28th,” she confirms. “Would customer and market insights that help mar- needed because it has learned what you like
you like to pay for that with your credit card keters sharpen their targeting, personalize and are most inclined to buy,” says an ana-
on file?” You step inside the store, bypass the customer engagements, design new prod- lyst. Amazon is so good at this that it’s even
long lines, and grab your order—no fuss, no ucts, and even craft better ads in real time. considering what it calls “predictive delivery,”
muss. Welcome to the world of artificial intel- Today’s machines are smart and ee- sending consumers stuff they haven’t even
ligence (AI). rily human. IBM’s Watson “is loquacious; it ordered yet. If customers don’t want it, they
This is just one example of how AI can tell jokes, answer questions, and write would just keep it for free. Although such de-
has exploded onto the marketing scene. songs,” notes one observer. “Google’s AI can liveries may still be a while off, Amazon uses
Starbucks has long been into cutting-edge now read lips better than a professional and such AI predictions to keep the right stock in
technology—a full 25 percent of its transac- can master video games within hours. MIT’s warehouses or even on trucks to support its
tions are already placed through its smart- AI can predict action on video two seconds ever-more-popular one-day or even one-hour
phone apps. But My Starbucks Barista is before it begins. Tesla’s AI powers [its] inno- delivery promise.
more than just an ordering app. It uses arti- vative self-driving car.” Hosts of retailers are employing AI to
ficial intelligence to create personalized cus- Companies like Amazon have mastered improve how they service and sell to their
tomer experiences and manage real-time AI, harnessing insights and interactions customers. For example, home improve-
customer interactions, based on everything that let it understand and serve customers. ment retailer Lowe’s is experimenting with
from customers’ past transactions and pref- Amazon’s Echo brings Alexa’s AI magic to LoweBots—five-foot something, fully mobile,
erences to demographics, store trends and nearly 50 million U.S. homes. Beyond serving AI-powered robots that roam stores helping
inventories, and local traffic and weather as a valet for duties such as adjusting house- customers. The LoweBots detect custom-
conditions. hold appliances, controlling music, keep- ers who might need assistance and engage
Artificial intelligence is sweeping the ing shopping lists, sending text messages, them through voice and touchscreens. The
world. It involves machines that think and and answering questions on about any AI robots tap store and external data to an-
learn in a way that looks and feels human subject, Echo and other similar AI devices, swers customer questions, offer solutions,
but with a lot more analytical capacity. The
engine behind the AI’s explosive growth is
big data. Raw data is flowing in from every-
where: customer transaction and interac-
tion data, web and social media data, news
and environmental data, and data from
more than 50 billion connected devices—
everything from consumer wearables and
GPS technology to household thermostats,
washing machines, and cars. Companies
need to make sense of all that data for their
brands and consumers.
The human mind simply can’t grapple
with today’s glut of big data. But machines
can. However, more than just collecting and
tabulating mountains of data, AI analyzes it
at lightning speed to gain deep insights and
apply them to accomplish designated tasks.
AI learns as it goes along—the more data
it ingests, the smarter and more accurate it Artificial intelligence: The My Starbucks Barista uses artificial intelligence to
gets. “AI is the planet we’re headed to,” says create personalized customer experiences and manage real-time customer
one AI expert. “Machine learning is the rocket interactions, based on everything from customers’ past transactions and
that’s going to get us there. And big data is preferences to local traffic and weather conditions.
the fuel.” Elias Stein Illustration
140 PART 2 | Understanding the Marketplace and Consumer Value
lead customers to merchandise in the store profile of an individual in literally millisec- billion locations every 15 minutes, letting
(or order online merchandise that’s not in onds,” explains an IBM executive. Watson munch on troves of data to gauge
stock). They even offer text and video tuto- Using such analytics, Watson can give mar- how weather affects consumers’ moods,
rials. Meanwhile, the LoweBots keep tabs keters precise, real-time views of customers health, and buying. It recently used a com-
on store-level data and analyze customer and put the insights it learns into action, using its bination of this weather data, consumer
shopping patterns. They’re “learning things AI powers for everything from data analysis and Google searches, and pollen counts to ad-
that we never knew before,” says a Lowe’s media planning to audience targeting and actual vise a medicine maker on which media to use
Innovation Labs manager, like “what is hap- content creation. According to one account: in various markets and when.
pening at 3 o’clock on a Tuesday” in any Despite all these remarkable applications,
As part of a Toyota campaign, for example,
given store. AI is still in its early stages. “We’re still in the
Watson became a copywriter, crafting mes-
AI does more than just serve custom- saging for the carmaker’s Mirai model based
dawn of AI adoption,” says a technology ex-
ers. It also helps marketing managers on [big data analysis of] tech and science pert. “It’s a new frontier and one that will re-
shape marketing strategies and tactics. For fans’ interests. Earlier this year, it trans- define the relationship between consumers
example, IBM has formed a new division formed into a doctor, promoting Theraflu and brands.” As an industry, AI will skyrocket
called Watson Advertising, built around while answering questions about various flu from current annual revenues of $8.1 billion
its AI supercomputer Watson, which first symptoms. For Campbell’s, Watson put on to more $105 billion worldwide by 2025.
gained public recognition when it bested its chef’s hat, personalizing recipes within And that doesn’t include the trillions of dol-
human contestants and won $1 million on display ads using data about consumers’ lars’ worth of retail sales that AI will facilitate.
locations and what ingredients they had
Jeopardy! Watson can ingest hundreds of “AI is going to be like electricity or the inter-
on hand. For a major partnership with H&R
millions of pages of data each second. IBM net,” says the Lowe’s technology manager.
Block, Watson turned into a tax expert, de-
has now turned Watson’s talents toward ploying an AI smart assistant to help clients
“It becomes so interwoven . . . it takes all of
marketing. For example, factoring in emo- find tax deductions.” this other stuff that we’ve been doing for so
tion, tone, language, sentiment, purchase long and it makes it better than the sum of
history, and social media interactions, IBM recently bought The Weather its parts.” Google’s CEO puts it more simply:
Watson “can generate a psycholinguistic Company, which produces forecasts for 2.2 “It’s more profound than fire or electricity.”33

or smartphone. Such systems allow managers to get the information they need directly and
quickly and tailor it to their own needs.

Author We finish this chapter by

Comment examining three special Other Marketing Information Considerations
marketing information topics. OBJECTIVE 4-5 Discuss the special issues some marketing researchers face,
including public policy and ethics issues.
This section discusses marketing information in two special contexts: marketing research
in small businesses and nonprofit organizations and international marketing research.
Then we look at public policy and ethics issues in marketing research.

Marketing Research in Small Businesses and Nonprofit

Organizations
Just like larger firms, small businesses and not-for-profit organizations need market infor-
mation and the customer insights that it can provide. However, large-scale research studies
are beyond the budgets of most small organizations. Still, many of the marketing research
techniques discussed in this chapter can be used by smaller organizations in a less formal
manner and at little or no expense.
Small businesses can obtain much useful market and customer insight without spend-
ing a lot of money. Consider Innocent Drinks:34
As friends and students at Cambridge University in the early 1990s, Adam Balon, Richard Reed,
and Jon Wright wanted to set up a business together. They believed that natural fruit smoothies
would be a great healthy pursuit and would make it easy for people to do themselves some good.
The lads began with very informal marketing research: instead of using a long questionnaire,
they simply had people at festivals test their first batch of smoothies and just asked them whether
they should quit their jobs to continue making these drinks. They had a simple bin that said “yes”
and a bin that said “no,” and at the end of the day, the “yes” bin was full of empty bottles.
All of them returned to work the next day and quit their jobs. They took their first delivery
of smoothies to 50 shops in London and offered them to the retailers for free; they only asked