Customer Equity Accelerator Podcast

Ep. 87 | Customer Targeting Gone Wrong: The Big Fish News Story


"Pause, take a moment and ask, should we target these people? What could go wrong?" - Allison Hartsoe


This week host Allison Hartsoe covers timely customer-centric news in the Accelerator. When Big Fish casino was named in a class action suit, it raised eyebrows. The social casino running on the Facebook platform was using data to aggressively target it’s customers. After some digging, this story of targeting gone wrong aired on PBS. How did this customer-centric data science strategy get so misguided? Allison Hartsoe recaps the PBS story (and one more) and interprets what it means for data scientists everywhere.     

Please help us spread the word about building your business’ customer equity through effective customer analytics. Rate and review the podcast on Apple Podcast, Stitcher, Google Play, Alexa’s TuneIn, iHeartRadio or Spotify. And do tell us what you think by writing Allison at or Thanks for listening! Tell a friend! See the full transcriptView all episodes.

Reference links:

If you want clearer insights from your data, check out how we can help:

Ambition Data Services

CLV Model Example

Marketing Reporting Example


Ep. 88 | Next Generation Data Warehouse (Part 1) with Claudia Imhoff Ep. 86 | The Business Roundtable’s New Focus

Show Transcript

Allison Hartsoe: 00:01 This is the Customer Equity Accelerator. If you are a marketing executive who wants to deliver bottom-line impact by identifying and connecting with revenue generating customers, then this is the show for you. I'm your host, Allison Hartsoe, CEO of Ambition Data. Each week I bring you the leaders behind the customer-centric revolution who share their expert advice. Are you ready to accelerate? Then let's go! Welcome everybody. This is Allison Hartsoe, host of the customer equity accelerator and CEO of ambition data. Recently there were two major news stories that you absolutely should not have missed if you care about customer-centric thinking. Last week I shared the first one, which was the business round table news and this is a group of policy shaping powerful CEOs who basically declared the purpose of a corporation was to no longer simply generate shareholder value but was to serve the various constituencies, including customers, listed first employees, vendors, and communities, major change in thinking.

Allison Hartsoe: 01:12 Of course, they were quickly called on the carpet by the B Corp Organization who said, Hey, we've been doing this all along, and it's not lip service. So catch last week's episode for all the gory details. This week I want to talk about another news story that's been growing more slowly over time but is really compelling. It's about target marketing gone radically wrong and a concept called protective classes, which we almost never hear about, almost never talk about. So let's dive in. In episode 80, I aired an interview with Aurelie Pols about building customer trust through privacy. And in the course of this episode we touched on, I swear it's a story that everybody must have heard by now, but it's a fairly famous story that people in analytics know about target and the outing of a pregnant teenager. Now I'll include a link to the story if you don't know it, but the summary is that it's a real story target used shopping behavior such as an increase in buying lotion and vitamin purchases to create a pregnancy prediction score.

Allison Hartsoe: 02:23 And they, you know, they did this similarly to what we do today where they looked at who created a baby registry and then they looked backwards in time at what products might predict that someone was going to create a baby registry. And that makes a lot of sense because their intent was to send coupons in order to corner the market on upcoming baby purchases. Not just wait until people were getting hammered with the fact that you'd had a baby and all these, you know, coupons started coming to you. And you know, for anyone who's had a child, you don't buy the crib when you have the baby, you do a lot of purchases beforehand, and you also kick off a lot of new purchases for things that you don't have in your house before you have the baby. So the thinking is right here. But to soften the stalker effect, they did something that was fairly clever and somewhat insidious.

Allison Hartsoe: 03:17 They threw in random coupons for wine glasses and lawnmowers and they also mailed to that person's neighbors so that it became a little bit less like they were targeting that particular person. And more that they were just kind of doing a general mailing of which it happened to be, oh, you know, coincidence, which is that nice serendipitous effect, you know, you look at a cruise website, and suddenly you get mailers about cruise ships and such. I know I've received that. Anyways, this worked and targets revenues grew and so Aurelie and I were discussing this privacy policy, and here is what she had to say about it.

Aurelie Pols: 03:56 I think the classical example, and I think some of my friends would typically say, well we're not sure this actually happened is you know the, the target's case, we heard many, many years ago about the data teams trying to define whether somebody was pregnant and that apparently I, I don't, I still don't know if the story's true, but the father of 15 years old was very, very upset with a target manager because he considered his daughter not to be pregnant. Besides, that's creepy aspect and things like that. I think what's really important to understand here as well is that by predicting a health state like pregnancy from your shopping behavior under the GDPR, you actually cross the line and I think a lot of data people work in data science don't understand that. If I use shopping behavior and I predict whether I prefer Banana Yogurt or strawberry yogurt, that doesn't really have a lot of consequences.

Allison Hartsoe: 05:02 And did you catch what she said by predicting a health state like pregnancy from your shopping behavior under the GDPR you actually crossed the line because your health state is protected data. Now that's obvious if you're a hospital, but direct to consumer marketers rarely think that way. We're always working so hard just to find customers that we don't sometimes stop and think about whether we should. And that is the main story I want to cover today, which is exactly about that target marketing gone wrong. The Big Fish Casino story starts with people's love of playing games on their phone. Oh, fairly innocent. And PBS news hour did a fantastic episode on this in August, which I recently caught and I will link to it in the show notes. In this episode, they interview a woman from Dallas, her name is Susie Kelly, and Susie basically sees an ad for the casino.

Allison Hartsoe: 06:04 She thinks it's a real place and she hears their tagline, play for free play for fun. And so she goes to the website, you know, figures out it's not real, downloads the app and then she can play poker, roulette, or slot machines. Now, this is not a person who at least in any story or any information I have found previously had a love of social casinos or went to casinos and any kind of regular basis. This is just like an average person that happened to download this app and start playing. So she starts playing with the set of free chips that they give her, but then she runs out, and of course in-app purchases start to take over. So you know, hey spend $5 and you can get x number of more chips. Well, in the first month, Kelly spends $8,000 buying chips that she knows she cannot turn back into cash.

Allison Hartsoe: 07:05 So at this point it's clear she's probably got an issue. She's got a problem, she's spending money and, and it's adding up little bits at a time, and she's not getting anything back in return. Well, nine months later, not only has she spent $40,000, but she can no longer deny that she has a gambling addiction and she needs to get out. So $8,000 painful, but $8,000 over several months. Very painful. So she emails the big fish customer service reps, and she says, please cancel my account. Please block my account. Please ban me from the site. She requests this from them nearly a dozen times, but the company never closes and never blocks her account. Instead, they do something that's really bad. They doubled down on her. They assign her a personal VIP rep who gave her free chips to keep her from leaving. Not only that, but the rep did this new kind of classic social selling where he starts checking in with her daily.

Allison Hartsoe: 08:17 He forms a friendship, he finds out what's going on in her life. So much so that when Kelly's mother passes away, they actually send flowers and more casino chips. So in total, she, over the time of this relationship, and they, they don't say exactly how long the relationship has gone on, but it's probably a year or two, she spent over $400,000. Now a real casino would be required to cut her off or face fines. So if she had walked into a real casino and then she said to them, I'm sorry, I need you to cut me off, I have a problem, I need to stop. Then they, they must comply, and that's it. And she walks away. They can ban her from playing. They can basically help her not hurt herself. But Online, this company isn't classified in the same way, so she can't walk away cause she's basically got the casino in her pocket on her phone the whole time.

Allison Hartsoe: 09:16 And the company is classified as an entertainment company, which allows them to escape gambling regulations. But here's where it's a little weird. These social casino games are not only more addictive than real casinos. In the PBS program, they say they're five times more addictive, but collectively they earn 5 billion a year, which is as much as the entire Las Vegas casino strip does. It's like if they walk like a duck and sound like a duck, it's probably a duck. They're earning as much as the Las Vegas casinos are. They're operating in the same way that the casinos are, but somehow they can't be held accountable, and that's a little bit of a gap. So clearly what they're executing is a high-value customer strategy. And PBS, in their episode, they turned up footage of Jose Bronson from Aristocrat and the gaming company, which is the gaming company that now owns big fish and Jose actually helped develop the VIP strategy.

Allison Hartsoe: 10:20 In the presentation, he cites that 3% of the customers will return 90% of the value to the company. Now that is a lot lower than most companies that we look at where the high-value customer base is more like 17% but also the purchases are maybe spread out a little bit more. So PBS then gets hold of some documents leaked from the actual program, and if you pause the screen and you look closely at the data, you'll see that they're tracking by individual using the Facebook id, they have age, sex and a graph of that person's spend over time showing that a whale, of course, a casino term, which for a high-value customer, a whale might be quickly ramping both in volume and frequency over an eight-month period. So it kind of looks like a shakedown. They get somebody who they can very quickly identify as a high-value customer, and then they hit them with a lot of marketing, and they hit them with this VIP person, and they basically feed.

Allison Hartsoe: 11:27 If they don't have an addiction, they might create an addiction, and if they do have an addiction, then they can easily wreck their lives. So they're tracking their day since the last visit. They're estimating their future spend. And so they can see prodo whales, people who look like they're going to become whales and sleepy whales, people who have decreased spend or frequency. So this is classic customer analytics, but this is where I think PBS also misses a nuance. Social casinos are bad when they prey on gambling addicts under the cover of entertainment. There's a secondary point here that the company is relying on an addicted health state, a protected class to generate nearly all of their revenues. And this is what I think is targeting gone horribly wrong. And the company that owns Big Fish aristocrat should really own up to this mistake, but instead, they've dragged their feet over four years and a class-action lawsuit.

Allison Hartsoe: 12:30 Facebook is the platform for these social casinos. They are obviously not innocent either. They know, again, PBS has footage of this in the episode. They know that their number one fastest-growing category on the platform is social casinos. They should have stepped up to raise a flag and say, Hey, what is going on here? These casinos are making so much money. And you know, it's really interesting that this is a fast-growing category for us. But why is it a fast-growing category? Did they ever stop to question? So not only did they let it slide, but they actually help the game developers target new players with better ads. Now again, if that's a general category, like no, like I'm buying t-shirts, or back to school clothes or whatever, it's less of an issue to target people with ads. But when it comes into a protected class is an issue.

Allison Hartsoe: 13:25 So what should you take away from these two stories about target marketing gone incredibly wrong. Let's always remember first that the people we target are not insights or outcomes, but real people. And they have a right to give consent and to be forgotten even when that's difficult to do even when that means that we lose money as a company. So for those of you working in data, I strongly feel it's our job to stop and think about the consequences. Ask yourself because of this, what will happen and run through a few scenarios imagining the different types of customers. See if you can uncover anything you wouldn't want splashed across the front page of the Wall Street Journal. I think that sunshine roll, which I've heard at other conferences previously as a really good rule of thumb. In other words, if we took this algorithm and we showed it to our customers, and they understood exactly what we were doing, would they be okay with that?

Allison Hartsoe: 14:26 Second, let's get a really clear understanding of protected classes. It's almost like data science ethics. California has a pretty strong definition of this, stronger than the US federal government does and they say simply this, you cannot target people solely on the following categories, sexual orientation, gender identity and gender expression, race, color, ancestry, national origin, religion, sex, including pregnancy, childbirth and related medical conditions, medical conditions, aids and HIV, disabilities, physical or mental, and I would put addictions in that mental disability, age 40 and older, especially genetic information, marital status, military or veteran status, political affiliation or activities and status as a victim of domestic violence, assault or stalking. Now, that's a lot of things. It's a lot of area, so what does it mean for us long term? I think that's up for debate. The laws are still being shaped as Aurelie and I talked about in that previous episode, we don't know exactly how all this will come through, but let's not be the kind of data scientists who play God, let's think ahead and protect others from playing God from wrecking lives and from generally being evil with data.

Allison Hartsoe: 15:54 As we learned last week, the desire to create shareholder value no longer trumps the customer. Let's be good to our customers and build strong relationships over time with their consent. If you want to talk more about the subject or perhaps find a way to become more customer-centric, you can always reach me at Allison at ambition data or at ahartsoe on Twitter or Allison Hartsoe on Linkedin. As always, links to everything we discussed including the four that are part of this program slash podcast thank you for joining me today. Remember when you use your data effectively, you can build customer equity. It is not magic. It's just a very specific journey that you can follow to get results.

Allison Hartsoe: 16:41 Thank you for joining today's show. This is your host, Allison Hartsoe, and I have two gifts for you. First, I've written a guide for the customer centric Cmo, which contains some of the best ideas from this podcast, and you can receive it right now. Simply text, ambitiondata, one word to, three, one, nine, nine, six, (31996) and after you get that white paper, you'll have the option for the second gift, which is to receive The Signal. Once a month. I put together a list of three to five things I've seen that represent customer equity signal not noise, and believe me, there's a lot of noise out there. Things I include could be smart tools. I've run across, articles I've shared cool statistics, or people and companies I think are making amazing progress as they build customer equity. I hope you enjoy the CMO guide and The Signal. See you next week on the Customer Equity Accelerator.


Key Concepts: Customer Lifetime Value, Marketing, Digital Data, Customer Centricity, Long-Term Customer Value, Marketing Leaders, Analytics, Creativity, Product Development, Audience Research

Who Should Listen: CAOs, CCOs, CSOs, CDOs, Digital Marketers, Business Analysts, C-suite professionals, Entrepreneurs, eCommerce, Data Scientists, Analysts, CMOs, Customer Insights Leaders, CX Analysts, Data Services Leaders, Data Insights Leaders, SVPs or VPs of Marketing or Digital Marketing, SVPs or VPs of Customer Success, Customer Advocates, Product Managers, Product Developers

Podcast Updates

Sign up to be notified when each week's episode is released.




Recommended Episodes


Ep. 62: Cultivating Customers with Prof. Elea Feit at Drexel



Ep. 54: 2019 Predictions Show Financial Impact from Marketing with Dan McCarthy



Ep. 43: The Customer Revolution is Here with Allison Hartsoe



Ep. 42: Intersection of CX and CLV with Ehow Chen


Listen to the Customer Equity Accelerator Podcast




Available on itunes  



Listen on Google Play Music