We are used to paying different prices for airline tickets, Uber rides, and
hotel rooms. But can you imagine a time when all sorts of retailers use data to tailor
their marketing and pricing for each individual customer? Welcome to "The Big Question," the
monthly video series from Chicago Booth Review. I'm Hal Weitzman, and with me to discuss the
issue is an expert panel.
Jean-Pierre Dubé is the Sigmund E. Edelstone Professor of Marketing
and director of the Kilts Center for Marketing at Chicago Booth, where he teaches an MBA
class on pricing strategies. And Jon Morris is the founder and CEO of Rise Interactive,
a digital marketing agency that came out of Chicago Booth's annual business-plan competition,
the New Venture Challenge. Panel, welcome to "The Big Question." Jon Morris, let me start
with you. Companies are collecting a huge amount of data these days. How are they using
those data to personalize their offerings? [Jon Morris:] Great question. They do it in
a few ways. The first is they're getting away from manual segmentation and actually--
[Hal Weitzman:] And explain what we mean by manual segmentation?
[Jon Morris:] Sure. So, rather than creating business
rules—we are going to market to just males, or just females—they are now able to gather
data on people on an individualized basis and learn about their attributes, and then
they can market to the individuals that have the highest propensity to actually be a customer or
purchase, as well as to customize the creative that each one of those individuals actually wants to see.
[Hal Weitzman:] What's the "creative," meaning . . .?
[Jon Morris:] A banner ad, an email, a direct-mail piece, their on-site experience, their
mobile-app experience, their social-media experience. There's a whole series of different
things that you can customize based on people's past purchase behavior, their online behavior,
their loyalty behavior, so that people actually start seeing creative that they're interested
in, and marketers are actually marketing to people who wanna see their creative, and spending
less money on people who don't wanna see their creative.
[Hal Weitzman:] OK, so what does that look like? You mentioned banner ads, something
pops up. We've all seen those kinds of things. It's directed at me. It seems almost like
a description of me. [Jon Morris:] Yeah. So, let's just say you
go to a sports retailer, and you are purchasing football equipment. The banner ads will know
what you've purchased in the past. They'll be able to analyze, based on the products you've
purchased in the past: these are the things that have the highest likelihood of being
purchased going forward, so they can show those products in the creative.
[Hal Weitzman:] And you work with lots of different kinds of companies. Who's doing
more of this kind of personalized marketing? [Jon Morris:] Retailers in general are doing
this fairly extensively, but we're seeing B2B companies doing this. We're seeing
hospital systems doing this. It really runs a very wide gamut in terms of multiple industries
that are doing this type of stuff. [Hal Weitzman:] J. P. Dubé, you're a
pricing expert. We were talking there about marketing. To what extent are companies using these data
to think about how to get prices right down to the individual or close to it?
[Jean-Pierre Dubé:] I think firms right now are still a little reluctant to genuinely
personalize in the same way that Jon described the personalization of, say, an email, or a
display ad, or some other communication. I think there's still a lot of misunderstanding
and disagreement about fairness, and when it comes to the prices charged, I think fairness
starts to play a really big role even though, as I said a moment ago, I think this is a
misunderstood concept in terms of how we define what's fair. That said, we've certainly got
industries where prices have been targeted and individualized for a long time. Airlines,
for example. We know that there's a lot of different ways in which dynamic pricing is
used in general. Dynamic pricing largely means that the firm allows its prices to evolve
over time based on the accumulation of data. So prices could evolve because a flight isn't
selling tickets—so demand reveals itself to be less pronounced than desired, so they
lower the price. A passenger, maybe through repeated interactions with the firm, reveals
their status as a high-willingness-to-pay person. They don't get access to the same
discounts. And if there's any doubts about that, all you need to do is log in to your
favorite airline's website, enter your loyalty card number or your loyalty membership, and
get your price quote. Then log out and do the same thing without providing the number, and
you often get very different fares. In fact, what the airlines will sometimes do is even
show you a different order in terms of the fares that are available. There's all sorts
of things they'll do to try and steer you toward different prices, including actually
charging you explicitly different prices for the same things.
[Hal Weitzman:] So, let's get into that issue about the controversy behind personalization.
Am I right in thinking, Jon, that in marketing this is less controversial, that I don't have
that reaction so much against having marketing personalized to me as I would about pricing
personalized to me? [Jon Morris:] Personalization is probably
the hottest trend in all of marketing right now. Every single company is trying to figure
out how they can make the experience more relevant and more meaningful for their customers
or their prospective customers. What they are all trying to do is collect massive amounts
of data. And the first thing they're trying to do is just understand who actually wants to see
their advertisements. Because if you think about it, when you spend money on marketing,
a percent of your money is always wasted on people who have zero interest in your product.
If you can ensure that you're spending it more on the people who are interested, you're
gonna get a higher return. Once you've actually identified those individuals, the next thing
is you're going to have to show them a creative. It could be a banner ad. It could be an email,
It could be a direct-mail piece. It could also be your web experience or your
mobile-app experience. And so, they are taking this data and customizing it and personalizing it. It might
be the products that they see, the offers that they get, the lifestyle imagery, the
language, the vernacular that is used in the language. There's a whole series of different
things that you can personalize once you have enough data on those individuals.
[Hal Weitzman:] Well, I guess what I'm asking is, do companies you work with get any pushback
from customers about the idea that their data are being stored and then mined to create this
advertising that you're talking about? [Jean-Pierre Dubé:] A lot of this depends
on transparency, right? I think the first thing is whether or not customers are capable of
determining that the messaging or pricing they received was somehow differentiated from
what others received, and this is not a new thing. Back in 2000, I think it was, Amazon
got into some trouble. There was a lot of press, there was a lot of discussion about
this online, when users determined that depending on which browser you were using, people weren't
seeing the same set of deals on top 10 CDs and top 10 books, so best sellers. So, when there's
some sort of differentiation in the kind of offers—and in particular, offers are being
made available to some but not to you—this is when you start to see a customer backlash.
But I think this really boils down to transparency. It also depends a little bit on culture. For
example, in Europe right now, and I'm sure, Jon, you're all too aware of GDPR, which
is a new set of rules and regulations that are gonna be implemented later this year regarding
how individual data can be used in the European Union. Number one, it's gonna set a set of
rules and regulations that'll be European Union–wide that all countries will have to
comply with. But among other things, one of the most controversial aspects of these laws is
that, in theory, it will give an individual user the right to challenge any kind of algorithmically
determined marketing communication to them, which could include a display ad but also
could include prices and other kinds of things, and that, in theory, this individual will be
allowed to challenge why they received that ad, which would then require the marketer
to tell them how the algorithm came up with the offer or the communication, which includes
going through the details of the algorithm, going through the exact data that were used—
what information did I use about you to come up with this offer? Whether or not this will
have any real bite in a court of law has yet to be determined. This is something we'll
need a judicial precedent in order to determine. But certainly right now it seems to limit
some of the scope for doing personalization in general. In the US, I think what you're
seeing is a lot more emphasis on the personalization, or at least segmentation, of prices that right
now people think being charged different amounts for the same thing is really unfair. It's
an odd thing because we've been charged different amounts for stuff for a long time. Some would
even question if it's legal in spite of the fact if you go to the movies, for the exact
same chair at the exact same showing of the same film, your age would determine what you
pay. There's a senior citizen's discount and there's a children's discount. So, there's
been price discrimination on age for at least as long as I've been going to the movies,
which, as you can tell by my hair color, is a lot longer than I wish. There's certainly a lot
of segmented pricing in the airlines, for example. So, it's funny that we accept these
things as par for the course in some segments, but when we hear about it in other settings
where we're not as used to seeing it, suddenly it seems unfair, and to some it might even
feel unethical or perhaps against the law. [Hal Weitzman:] Yeah. And why do you think
there is that tendency to feel that? We don't feel anything unfair about the airline
pricing, and, as you say, the senior discounts. And there's not people in the streets complaining
about those, and yet the idea that I would have a price personalized to me is something
that seems— [Jean-Pierre Dubé:] First, I doubt anyone would
say they think it's fair to have differential pricing in the airline industry. I think
people are kind of stuck with it.
[Hal Weitzman:] Yeah
[Jean-Pierre Dubé:] We may not be super inelastic in terms of our demand for a specific airline,
but we're definitely inelastic in terms of our needs to fly. There are jobs where I have
no choice but to fly. There are family occasions that require me to fly, which gives the airlines
a fair amount of power to implement these pricing regardless of whether or not I think
it's fair. In other industries like retail, for example, I might find it more unfair,
or I might be less willing to accept it because there may be more substitutes. But I think
a big part of this fairness debate stems from a lack of understanding about what is fair
and what's not. It seems the established wisdom now is that your layperson defines fair as:
we should all be charged exactly the same amount. That's fair. Here's the challenge
with that definition of fairness, and this is why I think more thought is required. I'm
sure we've all seen the movie "Minority Report." You've seen "Minority Report"? "Minority Report"
is a science-fiction film where a security force, using beings that have magical powers
and technology, purport to be able to anticipate individual's actions before they occur. And
the idea would be there to target and personalize a police action against someone based on a
predicted and anticipated behavior. Now, let's first review the facts. Number one, that's
science fiction. Number two, it still required individuals to have magic powers, these forensic
beings, so it wasn't really pure technology and data. But most importantly, even with this
almost deus ex machina, the magical beings, in spite of that, the climax of the film reveals
that the technology was imperfect. That's an important detail. Targeting is imperfect.
We can't get inside your mind and exactly know what you're thinking. We will always
make statistical errors and modeling errors when we're trying to predict behavior. So
now, let's step back from that. We're not doing perfect targeting; we're doing imperfect
targeting. What does the theory tell us, because we do have established economic theory on
what happens when a firm can engage in imperfect targeting. The answer is: it can go any way.
I know people don't like this "it could go any way; it depends," but the reality is when
a firm engages in imperfect targeting, where there's statistical error, classification
error, etc., the results are ambiguous. You can have scenarios where more consumers get
served as a result of targeting, and the consumer population could be made better off. If you
look at total value delivered, society could benefit from targeted pricing. Similarly from
targeted marketing. But it's also theoretically plausible that, with targeting, enough people
are charged higher prices or given disadvantageous— from their point of view—marketing that consumers
as a whole could be made worse off. Let's think about it. Let's imagine that a firm
engages in personalized pricing--this is probably the most contentious kind of targeted marketing--and
that more customers are served as a result, that the majority of customers actually are
targeted a lower price than would have been the price if everyone was required to be
charged the same amount, but in spite of that, a small minority of customers are paying more.
Is that fair or unfair? Well, let's think about the alternative. If we implemented a rule that said
by law, the firm has to charge everyone the same amount, fewer customers would get served,
so now we have to ask what's fair. Is it more fair to serve the majority, even if a small
minority gets charged higher prices as a result? Or alternatively, is it more fair to implement
a standardized, uniform price for everyone, where only a small fraction of the consumer
population can get served now, because prices are now too high?
[Hal Weitzman:] Yeah. A lot of interesting issues there. But there's still the social
taboo kind of exists. You make a very compelling argument.
[Jean-Pierre Dubé:] I just think people have a very imprecise definition of fairness. We
need to establish what is fair and equitable. Is it that everyone gets charged the same amount,
or is it that everyone gets access to the same value?
[Hal Weitzman:] In terms of the marketing side, and we talked about offers and coupons,
which of course are a way, aren't they, to personalize pricing while keeping the official
price the same for everyone. What is the relationship between the personalized marketing that you
do, or the segmented marketing that you do for customers, and the pricing strategies?
[Jon Morris:] In the groups that we generally deal with, oftentimes they're separate.
But one group in particular that we are noticing is the pricing strategies of
our clients that are selling commoditized products.
And what's interesting, especially as everyone is trying to figure out how to
compete with Amazon, is pricing becomes a critical component to what is the demand for their
sales, what is the impact that their marketing is going to have. And it's an interesting thing,
as we're talking not just about fairness but just overall pricing strategy, in the online
world, Amazon changes all the rules. Oftentimes they have a huge pricing advantage, and people
are really trying to focus on the personalized aspect so that if they can customize offers
or they can customize pricing to individuals, it gives them a chance to potentially compete
with the 800-lb. gorilla in the retail world.
[Hal Weitzman:] J. P. Dubé, because of some
of the hesitations that we've talked about, companies are not, am I right in saying,
companies are not using targeted pricing in the same way that they're using targeted marketing?
Is that fair? [Jean-Pierre Dube:] I would say it's just
not as pervasive. There are, of course exceptions, there are many retailers, not just in
the digital environment but in the physical store environment, that are starting to experiment
with targeted coupons, so there would be a regular price on the shelf and then targeting
or personalization of prices would entail getting a special promotional offer for some
discount off the regular price. But that is just price discrimination. Your highest-willingness-
to-pay customer segments don't get an offer, so they pay the shelf price. That becomes
the high price. And then successively lower-willingness-to-pay segments get successively
better coupons to get deals or discounts off of that regular price.
[Hal Weitzman:] Is that the way to sidestep this issue of whether it's fair or not?
[Jean-Pierre Dubé:] Well, it could still be perceived as unfair. Again, I still view fairness
as a subjective thing because we don't have a legal opinion or precedent on what is fair.
But if people found out that some customers got coupons and they didn't, this could be
perceived as being unfair. [Hal Weitzman:] Right. Has there been a backlash
against coupons? [Jean-Pierre Dubé:] Absolutely. First of all,
there's been in the media, the "Atlantic," for example, last year had a pretty scathing article
about the potential scope for unfairness and abuse of targeted marketing, especially prices.
The Council of Economic Advisers actually wrote an entire report in 2014 on targeted
marketing with big data, and then the following year, in 2015, wrote a second report that was
explicitly about differential pricing and big data. If we were to take that report at
face value, their prediction was it would be extremely unfair. The way they viewed it
as unfair—and I have to admit, the report mischaracterizes economic theory. I don't
think the report was carefully done—but suggests that differential pricing would expropriate
value from consumers and would then transfer that to shareholders of firms. And as I already
indicated earlier, that's not actually a theoretical per se outcome; that is theoretically plausible
for consumers as a whole to benefit from targeted pricing.
[Hal Weitzman:] OK. Sorry, go ahead, Jon.
[Jon Morris:] From what we see from the marketing side, very rarely are we having conversations
with our clients as we're talking about offers as it relates to fairness. It is more in a
conversation of: What can we do to maximize average order value? What can we do to maximize
the conversion rate? But I'll give you just some interesting nuance. If you go to the
shopping-cart page and you're ready to check out, and there's a little line that says put
in your coupon code to get a discount, oftentimes your conversion rate will drop, meaning
that the percentage of people who actually go to purchase decreases. Couple of reasons.
A: some of the people actually leave to go see if they can find a coupon code and they
miss that, and other people, as it relates to this feeling of fairness, feel like, hey,
someone else is getting something that I'm not, and so there's a negative emotional impact
that actually loses sales. It's interesting: there's some aspects that drive sales, but
in some aspects, it actually decreases the number of people who are willing to purchase because
of what you put onto that thank you page, rather that conversion page.
[Jean-Pierre Dubé:] And scientifically, this is actually an old phenomenon. Our resident
behavioral economists here at Booth would probably be eager to talk about the literature on transaction
utility. Economists tend to focus on the acquisition utility of products and services. Theoretically
I think of this as the consumption benefits, but it can be more. Transaction utility literally
indicates utility from the merits of the deal. The fact that I think I got a deal gives me
intrinsic utility. It's maybe psychological and perceptual, but as Jon's example indicates,
it can affect choice. The fact that I don't have a coupon but there's a slot for it that
indicates someone else might, might suddenly make me feel I'm getting less transaction
utility and paradoxically could affect the sale.
[Hal Weitzman:] Right. Which, it does seem like a fairness issue, no?
[Jon Morris:] From the consumer standpoint, they feel that if they're the ones that don't
have that coupon that they are being discriminated against in a negative way, and it impacts
oftentimes— [Hal Weitzman:] On the other hand, as you
say, it could be a huge opportunity, because even if you give someone 25 cents off a $300
transaction, they might feel like— [Jon Morris:] They feel like they won something.
[Hal Weitzman:] —they've got a good deal. When we're talking about personalization,
give us a sense—from your, you deal mainly with the marketing side--how personalized
is it? Is it really down to the individual? Presumably—you talk about creative—you can't make banner
ads for each person, so how personalized would be the banner ads that people see?
[Jon Morris:] It is actually going down to
the individual. There's a technical term called device stitching.
[Hal Weitzman:] Device stitching? [Jon Morris:] Device stitching. If you think
of an individual, they have a physical address, they have an email address, they have a
mobile-phone number, but they also have a few other things. They have a cookie ID. On their computer,
there's a cookie that is associated with them. They also have a device ID on their mobile
phone. Device stitching is trying to take all those different pieces and bring it so
that you are marketing to one individual. Rather than marketing to someone's laptop
and marketing to their mobile phone, you are marketing to an individual that happens to
have multiple devices. That's the first key element. Once you have that, then you want
to start layering in different data, such as: What is their transaction history? What have
they purchased in the past? You can put tags on websites, where you can literally monitor
every single thing a person does on a website. What sections did they read? What did they
put in the shopping cart but didn't purchase? What did they actually purchase? And so you
use that, and then you're getting into artificial intelligence, where the artificial intelligence
is determining, OK, for this creative, it's a template where there's different things
that you can . . . you know, the background image can change. The language can change. The product
you show can change. It's basically allowing you to do things at a scale that you've never
been able to do before. Because, if you have to create a creative for every single individual
and you have millions of customers, there's not enough money and time to be able to do
that in a cost-effective manner. But now, every single individual can get a completely customized
creative that is based on their attributes and what's unique to them.
[Hal Weitzman:] What proportion of companies do you think are really doing this to its
full extent now? [Jon Morris:] Based on a study I saw last
year, the entire market cap of personalized marketing is around $640 million, which is
tiny. The expectation by 2025 is $35 billion. What we're seeing now is clients are going
from talking about it to starting to earmark tens of millions of dollars to do personalization
in a major way. The hardest part, and the reason why it is in its infancy, is that companies
did not build their organizations and their data and their infrastructure with this concept
of doing personalized marketing, so all their data are in very siloed, disparate places
that oftentimes are hard to export, are hard to merge together. What's happening right
now is more in the way of investments in the data infrastructure.
How do we make it so that we can get our CRM
data, which is where all the "people" information is, with all the transaction data, with all
the loyalty data— [Hal Weitzman:] You mean those are housed
typically in different departments. [Jon Morris:] Very different departments.
Oftentimes there's political reasons, where someone doesn't want to give access to this
data and other people do want to give access to the data. But then you've gotta bring all
that data together. Once you have that data together, you have a major competitive advantage
because that's when you can start getting into personalized marketing at a scale like
no one else can. [Hal Weitzman:] OK, so it sounds like all
we have to do to make the revolution happen is to break down all the silos between the
deptarments, which should be very easy, right? In terms of—
[Jean-Pierre Dubé:] It's not easy. [Hal Weitzman:] No, no, I'm being sarcastic.
[Jean-Pierre Dubé:] No, but yeah, it's actually amazing how many firms have all the resources they
need to be doing this, but this is really an org-design problem.
[Jon Morris:] Yep. [Hal Weitzman:] Right. It almost sounds like
the old marketing department does not work. The new marketing department is something
that has to link all the other departments together.
[Jon Morris:] One of the things that . . . we're talking to a major retailer right now, where
they're literally going through this transformation. They've earmarked tens of millions of
dollars for personalization. They are reducing budget across multiple groups and putting
a series of people that are forced to work together to build the infrastructure to create
this data. And what you'll find is publicly traded companies have a much harder time because
the organizational infrastructure is so built up over time, compared to PE-backed organizations
that have, I say, more nimbleness and the ability to bring this data together. Those
are some— [Hal Weitzman:] Presumably the older companies
have had it harder than newer companies. [Jon Morris:] Yep.
[Jean-Pierre Dubé:] Just take retail. If you compare Amazon to pick your favorite national
supermarket chain—as we know Amazon's now moving into the grocery space—Amazon has
hundreds of data scientists. They're hiring our PhD students now to come in and work with
their data. Amazon's infrastructure, their IT architecture is built around being able
to use data and make decisions on data. The surprising part of this is some of the stuff
that Amazon's doing—let's ignore the fact they have a digital platform, but just the
database stuff they're doing—most large supermarket chains could have been doing since the '90s.
Chains have been collecting individual customer data using loyalty cards for decades now.
The human capital required exists to implement and use these data, and implement targeting has
been around. In PhD programs, we've been teaching students how to do personalized pricing and
personalized marketing since the '90s. We have faculty here at Booth who wrote papers using
grocery-store loyalty data to design personalized-pricing strategies where every single customer
ID would have their own personalized price. So the methodology has existed. The understanding
of how to use data has existed for the better part of 20 years,
maybe 30 years. And the types of data you would need have been available to firms. This
is literally a constraint based on who's working for the firm—does somebody actually work
for the company that would know what to do with the data if they existed? Having the
data set up in a way that they're accessible and implementable. And then, of course, a volition
by the firm to implement these kinds of strategies in the first place.
[Hal Weitzman:] So, the same kind of things that are holding back completely personalized
marketing are holding back the personalized pricing?
[Jean-Pierre Dubé:] There's obviously the fairness issue with pricing, which is a deterrent
for a lot firms to engage in it; but I think before even a firm would be ready to engage
in personalization, even if they wanted to do this, you would need to rethink how data
are shared, who's in charge of marketing— do you actually have people in marketing who
would now how to design and implement these strategies. There's just a lot of knowledge
and awareness that's missing. [Hal Weitzman:] So, in those 20 years, we haven't
made huge progress toward realizing the potential of these data that companies have been collecting.
[Jean-Pierre Dubé:] We should say 30 years because— [Hal Weitzman:] 30 years.
[Jean-Pierre Dubé:] Yes, 30 years. I'm thinking of papers that would seem very innovative
to a practitioner right now that were published in the early 1990s on exactly this topic.
[Hal Weitzman:] I was gonna ask, the next 30 years, is that when the barriers are gonna
be broken down, and we'll actually get to personalized marketing and personalized pricing?
[Jean-Pierre Dubé:] What's happening, and this is largely in the digital space, Silicon
Valley in particular, is that more and more firms are now hiring a blend of personnel
that includes PhD-trained individuals with PhDs in economics, PhDs in marketing. They're
hiring actual tenured faculty in economics and in business schools to come in and be
chief economists. And I think the companies that have been doing that kind of change in
their personnel are also the ones that are engaging the most actively in these kinds of practices.
[Hal Weitzman:] OK. Jon Morris, what do you think is the future for personalized marketing
over the next few decades? [Jon Morris:] Couple of things. One, you're
starting to see VPs or SVPs of personalization. One of the things I personally believe is
you can't have a goal unless you put a budget against it.
[Hal Weitzman:] So, that's what J. P. was talking about, bringing in someone specifically
to bring all the data together. [Jon Morris:] Yep. There's actually budget
finally getting allocated to do this. But I think the biggest difference between the
past and the present is that the ability to leverage this data now is in a better place
than it has ever been. There also is a bunch of advertising technology companies out there
that allow you to leverage this data in a faster and more efficient manner than the
past. I believe that we are at that turning point where you're going to see--we're already
seeing it with our customers--that the number of customers willing to invest in this is
growing and growing. One great example is I've seen a multibillion dollar company that
created a start-up within its company to really sell pretty much the exact same product or
service, but because they wanted them to get away from the old infrastructure and the old
technology and be able to build from the ground up, they felt that it had to be outside their
walls. They built the entire data infrastructure that was necessary, and they did it at a fraction
of the cost because these legacy systems have layers and layers of different pieces of technology
on it that it's just too costly to manipulate, and so oftentimes you have to start over.
[Hal Weitzman:] OK. And Jean Pierre Dubé, you spoke persuasively against the idea that
this is unfair, but do you see the concerns about unfairness or about privacy, perhaps,
for data going away, ebbing away over the coming decades?
[Jean-Pierre Dubé:] It's a lack of transparency. When I go to the movies, I know that I was
charged a high price because I'm a middle-aged adult and not a child or a senior, at
least not yet. When I go on the airlines, I have a slightly--it's still vague--but a
slight idea that I'm paying a higher price because it's a week until my flight and the
airline tickets get more expensive if you wait too long. In the digital domain, I think
the concern for a lot of people is just that lack of transparency. I got served up an offer.
My offer wasn't all that appealing. I find out that there's many other offers that were
much better, and I can't understand why I didn't get the better offer. And even worse
is this idea that the firm or the marketer is tracking what I'm doing, and somehow my
interactions with the firm in the past are gonna lead to less favorable interactions
in the future. So, in some sense, I'm worried: Am I gonna get punished for my loyalty? That,
of course, is an interesting question. Do we think that as customers become more attuned
to getting targeted offers, will they start altering their behavior to try and get better
offers? I'll give you an example. I ran some targeted promotional campaigns in China with
a telecom company, and the targeting scheme was meant to illustrate the new opportunities
for data. It wasn't really a big-data project, but it was about showing what is data. In
this particular case, we targeted based on real-time location. We used the GPRS signal
as an indicator of where you were in real-time, and then your proximity to the movie theater
was then deemed to be relevant to whether or not you're a prospective customer or not.
So, we targeted prices accordingly. And one of the concerns from the marketing folks on
this campaign was: over time, if people who are really far away from a movie theater are
getting systematically better offers on their SMS, does that mean that people might actually
stop dwelling in malls? Will I start dwelling somewhere else waiting for my offer and then
I'll go to the mall once I've received my offer? Once you do that, you're starting to
unwind the targeting scheme, and this brings us back to an important theoretical point that
targeted marketing works under the assumption that there's no arbitrage. As soon as customers
can decompose what it is you've done to target and are willing to change their behavior to
unwind the fences, targeting might not be so successful anymore.
[Hal Weitzman:] When we get to that point, we can sell personalized arbitrage software
for everyone to try and get the best coupon offer.
[Jean-Pierre Dubé:] Which happens. [Hal Weitzman:] Unfortunately, on that—we'll
have to come back and talk about it next time because unfortunately at the moment our time is up.
My thanks to our panel, Jean Pierre Dubé and Jon Morris. For more research, analysis,
and commentary, visit us online at Review.ChicagoBooth.edu,
and join us again next time for another "The Big Question."
Goodbye.
For more infomation >> LSU says there is no investigation into WIll Wade - Duration: 6:48. 
For more infomation >> Monday's storm is not the only one in the forecast - Duration: 3:25. 
Không có nhận xét nào:
Đăng nhận xét