The iPhone of breast pumps and behavioral ad targeting
|Jun 28||Public post|| 1|
Ranjan here. Today we talk about targeted advertising.
Willow and Elvie
Midway through the pregnancy of our 2nd child, my wife asked me what I thought of a device called the Willow Breast Pump. While I had exhaustively researched every possible baby gadget for our first kid, this was something I hadn't really explored.
Willow, and a similar product from a company named Elvie, were both iPhone-ifying an incredibly staid industry. I won't get into the details for the uninitiated, but the world of breast pumping is one of those things, before parenthood, you just have no idea (or at least I had no idea). As the pressure on mothers to exclusively breastfeed has increased, the technology enabling them to do so feels like its straight out of a 1980s hospital. It’s a constant, audible, omnipresent reminder of your wife’s sacrifice.
These new disruptive products were quiet, discrete, and comfortable - the exact opposite of what's out there now. If startups are meant to solve painful, awful problems, this was certainly one worth attacking.
The reviews were fairy consistent. These products were good, but early.
Willow was close to launching it's 2nd gen product at the time, and the Elvie was soon-to-be launching (they're both out now). The consensus seemed to be, the Willow Pump had a steep learning curve, which often required a (complimentary) online coaching session, wasn't quite as effective as traditional pumps, and was multiples more expensive than current products. In other words, it sounded like any early-stage tech product. There was a ton of promise, it would work pretty well, but there would be a lot of frustration.
Now, I love nothing more than buying the first generation of a gadget and poking around. My wife is supportive of my early-adopting addiction, but usually at an arm's length. Even after walking her through the universally acknowledged downsides, she was adamant about wanting to try it. I was kind of excited as it meant a cool new gadget in the house.
Then I saw her Instagram feed.
It was plastered with sponsored Willow posts. Elvie ones too, and that product wasn't even out at the time.
Logical and Targeted
This annoyed me. But it made a lot of sense. She follows a bunch of Mom-related Instagram accounts. There were baby bump pics posted. Instagram could probably even figure out our approximate due date from photo captions. Any old algorithm knew she was an expectant mother.
Our son was born. The Instagram ads continued. We bought the Willow. It's pretty incredible.
It was exactly as expected. It's revolutionary. It's early. It's iPhone-ish. There have been frustrations. It was expensive. There is no doubt in a decade this style of pump will be better and everywhere. Investors, get in on their next round.
But I just can't shake that discomfort with the marketing process. Both of these companies have raised a ton of venture (Willow $42.5mm, Elvie $42mm) which naturally meant they’d be hitting the gas on platform growth strategies. Repeatedly promising to solve one of the most difficult and painful parts of an entirely difficult and painful time, in a hyper-targeted, algorithmic manner. Is that okay?
Where do we draw the line with micro-targeted advertising?
This was the perfect grey area case study. Perfect enough that I'm willing to risk my wife's ire after she sees I posted this (you’re the best wife and mother ever!).
Willow is trying to tell their target customer about their cool new product. It’s just a standard marketing problem to solve.
Sitting poolside. A number of Willow-clad women doctors. A mommy-blogger and her luggage, jet-set ready. Willow pumps on the beach. Netflix and pump. You can be successful, beautiful and not give up an ounce of your lifestyle. You can be free. All thanks to the Willow Pump. It's so good.
But what's an ethical standard for how you target your consumer? You'd naturally want to target expectant or recent mothers.
Should there be a limit of 'x' posts per day? Her feeds were littered with repeated ads.
What about the timing? Can the platform ascertain the approximate due date? Can you pay a data broker for 3rd party info that confirms this (from babycenter.com or something)? Can you step up the number of ads shortly after birth, knowing how sleep-deprived the customer will be?
Can you target ads between 1am-5am, knowing those are the hours a mother is exhaustedly breastfeeding while scrolling Instagram to stay awake?
What is the line when you have incredibly clear data points as to when your target customer is at their most susceptible? What if you have a great product that you genuinely believe will solve their problems; do the ends justify the algorithmic means?
What is legal?
Full disclosure: my legal knowledge base on this topic is from just a few web searches.
Last week we discussed how the subscription-ification of the economy will require a rethink of subscription management processes. The evolving nature of targeted advertising requires an even more urgent discussion around what is okay?
The most recent regulatory development I encountered started with the 2016 ProPublica story where they showed how real estate ads could systematically discriminate against specific racial groups. In March 2019, Facebook announced they would prevent housing, employment and credit services from excluding audiences using gender, zip code or race.
But these are all focused on the exclusionary, not the inclusionary. What about limits on who, when and how advertisers can target someone?
Start with the kids
There was a Do Not Track Kids Act in 2011 that was expanded in 2018, which requires parental consent for data collection on minors as old as 16. It prevents them from using any collected data to behaviorally target kids. (Note: It freaked me out to read that 90% of children under two have an identifiable digital footprint, that can begin even before they are born. Our kids probably fall into this). I would hope, societally, we can agree that kids under 16 should not be tracked and targeted. Protecting kids seems like a good place to start.
But what about a woman who is in her 3rd trimester or has just given birth? What about a father with a new child? After the birth of our first kid, anyone promising to make my kid a little safer and my life a little easier had my undivided attention. The baby industrial complex is strong and I was incredibly commercially vulnerable.
Targeting someone who has been identified as overweight with a diet product seems reasonable. But what about seeking out those who have been posting depressive things? Sentiment analysis is only getting better. Marketing payday-ish lending products to lower-income consumers is uncomfortable, but it makes sense. You don’t want to waste ad dollars on wealthy consumers. But what about targeting them a few days before the typical paycheck cycle, when you know they're at their most financially strapped (Earnin app, anyone?)?
Spectrums and Slippery Slopes
“People consistently tell us that if they’re going to see ads, they want them to be relevant. That means we need to understand their interests.” - Zuck in the WSJ
Writing this post reminds me of how difficult regulating this will be. Everything I’ve described above is theoretically in the consumer’s interest. Moreover, so much of this stuff exists on a sliding scale or spectrum, and advertising has always been targeted in some capacity. I’m currently at my parents house and every CNN ad break is stuffed with pitches for reverse mortgages and prescription medications.
But it's only a matter of time before we move past broad demographic cable TV and radio ad buys to personalized, targeted ads lining all of our media consumption. As consumers, we're only going to continue to create more identifiable data, and the technology which can comprehend and process that data this data will only get better. How do we begin to set limits on what is acceptable?
I always lean towards transparency as the nearest-term solution. Providing consumers with a greater understanding of how and when they’re targeted is a good start. The Facebook Ads library, can show me what ads Willow posted, but not how much they spent or when and how they targeted us. I would love to know that.
Finally, as this piece is on parenting + tech, I’ll leave you with this photo that I hope does not create any digital footprint for my new kid. This is how this piece is being typed:
What I’m Reading
…or more accurately, what I’m listening to. I’m a big fan of Axios’s Pro Rata podcast hosted by Dan Primack. Relevant to this piece, the podcast has recently been sponsored by Facebook. The ad segments are Facebook espousing the benefits of….targeted advertising. I’ve seen Axios get a lot of flack for having big corporate advertisers, but it’s very comforting to me to see them sponsored by Facebook, in the very same pieces of media where they rip apart Facebook’s business model.
It kind of has an pre-digital media feel to it (the Facebook ad starts around the 1:56 mark):
Black Market T-Mobile Location Data Tied to Spot of a Triple Murder: From the data brokering frontlines, an insane story from VICE:
In this Texas case, the data of the bounty hunter's phone first trickled down from a telecommunications company acquired by T-Mobile, according to a screenshot of the phone ping obtained by Motherboard. Then, the data moved through a location aggregator called LocationSmart. From there, LocationSmart sold that access to CerCareOne, a company that operated in secrecy, and which told its clients to not even reveal the company's existence.
CerCareOne had around 250 bounty hunter and bail bondsman clients for real-time location data, according to previously leaked CerCareOne documents.