Facebook’s Data Problem (Thinks Out Loud Episode 202)
Looking to drive results for your business? Click here to learn more.
Facebook’s Data Problem (Thinks Out Loud Episode 202) – Headlines and Show Notes
- Facebook Ad Reach Claims Don’t Match Census Data, Pivotal Notes | Fortune.com
- Facebook Responded to the Backlash Over Its Ad Metrics | Fortune.com
- Facebook Figured Out My Family Secrets, And It Won’t Tell Me How
- Panopticon – Wikipedia
- Procter & Gamble Cut Up to $140 Million in Digital Ad Spending Because of Brand Safety Concerns – Adweek
- The CMO Report — September 5, 2017 | The Full Monty
- P&G Cuts More Than $100 Million in ‘Largely Ineffective’ Digital Ads – WSJ
- 7 Quick E-commerce and Digital Strategy Insights You Won’t Want to Miss: E-commerce Link Digest
- Facebook now has 2 billion monthly users… and responsibility | TechCrunch
- Facebook introduces Watch, its redesigned tab for video programming – The Verge
- Social Advertising Isn’t Really Driving Conversions – eMarketer
- 6 Proven Digital Marketing Trends (The Best of Thinks Out Loud – Episode 200)
- BarnRaisers15 eCommerce case studies show big results from small changes
- 5 High-Impact E-commerce, Digital and Social Strategy Posts You Won’t Want to Miss: E-commerce Link Digest
Subscribe to Thinks Out Loud
Contact information for the podcast: firstname.lastname@example.org
Past Insights from Tim Peter Thinks
You might also want to check out these slides I had the pleasure of presenting recently about the key trends shaping marketing in the next year. Here are the slides for your reference:
Technical Details for Thinks Out Loud
Recorded using a Heil Sound PR 30 Large Diaphragm Multipurpose Dynamic Microphone through a Cloud Microphones CL-1 Cloudlifter Mic Activator and a Mackie Onyx Blackjack USB recording interface into Logic Express 9 for the Mac.
Running time: 15m 32s
You can subscribe to Thinks Out Loud in iTunes [iTunes link], the Google Play Store, via our dedicated podcast RSS feed )(or sign up for our free newsletter). You can also download/listen to the podcast here on Thinks using the player at the top of this page.
Facebook’s Data Problem – Thinks Out Loud Episode 202 (September 6, 2017) Transcript
Well hello again everyone and welcome back to Thinks Out Loud, your source for all the digital marketing expertise your business needs. My name is Tim Peter, today is Wednesday, September 6, 2017 and this is episode 202 of the big show. Thank you once again for tuning in, I really do appreciate it. I’ve got a terrific show for you this week.
Now I have to do something a little unusual today and that is provide a disclosure: I’m going to be talking a lot about Facebook and I think it’s important to note that I own a small number of shares of Facebook, that represent about 0.7 percent of my portfolio. I don’t think that introduces any bias into the conversation, but, obviously, you should know that and judge accordingly. Also, of course, I think it’s really important to note that I’m about to look at the company from the point of view of marketing professionals, so please don’t take any of this as investment advice. This is all about what you do as a marketer, not as an investor.
Now with all that out of the way, it's really interesting what's happening with Facebook these days. They really represent a study in contrasts for marketers and customers alike and a lot of that contrast really has to do with their data problems. And they have a number of them.
Facebook’s Data Problem: Miscounting Subscribers
According to Fortune and Pivotal Research Group:
…Pivotal Research Group senior analyst Brian Wieser pointed out a large discrepancy between U.S. census data and the potential reach that the social network promises advertisers.
I’m still quoting the Fortune article here:
On Tuesday, Wieser issued a note pointing out that Facebook's Adverts Manager tool promises a potential reach of 41 million 18-24 year-olds in the U.S., while recent census data said there only 31 million people living in the U.S. within that age range.
For 25-34 year-olds, Facebook claims a potential reach of 60 million, versus the 45 million people counted in the census last year.
So to recap that, Facebook claims it’s reaching somewhere between 25 and 35% more 18 to 34-year-olds then the U.S. Census shows exist. That’s not great.
Now, there could be multiple causes for this:
- Some people have more than one account. Whether for legitimate reasons or less legitimate reasons (more on that in a moment), it’s entirely possible people have more than one account on Facebook and so the social network is double counting those people
- Older people (and, who are we kidding, it’s probably creepy old guys mostly) with only one account might be lying about their age for any number of reasons that I'm not getting into, right, or, most likely…
- Some combination of the first two
Still, that is an issue.
Additionally, a year ago, Facebook had to apologize for “artificially inflating” the number of video views on the site. Those are big deals, right? Those are really big deals. But they’re tough to reconcile with the fact that we also think that Facebook knows everything.
Facebook’s Data Problem: They Know Everything
At the same time, Facebook — with the possible exception of Google — has data about that’s second to none. Kashmir Hill wrote a piece on Gizmodo called “Facebook Figured Out My Family Secrets, And It Won’t Tell Me How.” It’s a crazy story, and I have to quote it at some length for you to get a sense of exactly how crazy. So, here’s the quote:
“And then there was Rebecca Porter. She showed up on the list after about a month: an older woman, living in Ohio, with whom I had no Facebook friends in common. I did not recognize her, but her last name was familiar. My biological grandfather is a man I’ve never met, with the last name Porter, who abandoned my father when he was a baby. My father was adopted by a man whose last name was Hill, and he didn’t find out about his biological father until adulthood.
The Porter family lived in Ohio. Growing up half a country away, in Florida, I’d known these blood relatives were out there, but there was no reason to think I would ever meet them.
I’m going to fast-forward through this…
I sent the woman a Facebook message explaining the situation and asking if she was related to my biological grandfather.
“Yes,” she wrote back.
Rebecca Porter, we discovered, is my great aunt, by marriage. She is married to my biological grandfather’s brother; she met him 35 years ago, the year after I was born.
Facebook knew my family tree better than I did.
Now that’s crazy. It's connecting people who, apart from this biological relationship several generations back, have nothing in common in the real world, and knew that these two people should know one another. I had a similar thing happen when Facebook recommended I add as a friend a contractor who’s done some regular work on my house, who I found online about 10 years ago, and have no connection to in any other way. Again, crazy how much they know about us.
Facebook’s Data Problem: Analysis
So, which is it? Is Facebook the panopticon, all-seeing, all-knowing, trapping us in its omniscient vision? Or is it a flawed system that defrauds advertisers and fleeces us using fake data?
The easy answer, of course, is “why not both?” But that’s not really fair to Facebook. And it’s not accurate.
I don’t think Facebook willfully lies to people. I strongly suspect, like many companies, they may view data through whatever prism is most aligned with their incentives. They clearly need to attract advertising dollars to drive their continued growth and where they have the greatest opportunities and incentives, they’re re going to position those numbers in whatever way makes those numbers most attractive to advertisers. Which might sound like they're lying, but I think it's more looking at the world through rose-colored glasses. They may not be saying things that are hundred percent accurate but I do think it is something that they’re incented to see most positively.
I also strongly suspect that they’re going to struggle to find those ad dollars in the future without seriously changing their product. Now they're doing some of this will already, which talk about in a moment, but I mentioned a few weeks ago that P&G cut $140 million from its digital advertising spend due to “brand safety concerns and ineffective ads.” P&G is the biggest advertiser in the world, spending $2.4 billion last year according to Advertising Week. Now, when Facebook cut its spend, the Wall Street Journal reported that “P&G indicated it wouldn’t pull back on overall Facebook spending.”
But, let’s be fair, they haven’t indicated whether they’re increasing that spend either. And this is where we get to the crux of the issue and how I would look at it if I were you.
Facebook’s Data Problem – Reach vs. Targeting
P&G has a very specific need: They want reach. That’s why TV advertising still works brilliantly for the company; they can put their message in front of more people on TV than just about anywhere else. And that’s why Facebook is itself getting heavily into the video space to compete not only with YouTube, but with traditional television as well. Its Facebook Watch offering is only the latest step in that direction.
But Facebook’s greatest strength right now — despite its 2 billion monthly active users — is not reach. It’s the ability to target those users individually. Even the “bogus” numbers of “how many 18-34 year olds are there on the site” aren’t terribly important in that context. Now don’t get me wrong; I’m not saying it’s great that Facebook’s numbers are off. It’s just not as big a deal as the Fortune article would have you believe… unless you’re trying for the broadest possible reach.
If you’re looking to connect with a tightly-targeted audience, especially at the top of the funnel, nothing compares with Facebook. Facebook knows more about your customers than just about anyone, including, possibly, Google. And marketing on Facebook works brilliantly. According to recent research from CPC Strategy and available on eMarketer,
…over one-quarter of US internet users had made a purchase after clicking on an ad they came across on Facebook. Meanwhile, just 7% of those who didn’t click on an ad completed a purchase as well. The survey also found that those who did click on a Facebook ad were roughly 3.5 more times likely to buy a product than those who didn’t.
Those numbers were dramatically higher for Facebook than other social networks. The same eMarketer article shows that “…just 1% of respondents ages 13 and older said they made a purchase based on a Snapchat ad they saw, and only 4% said they bought anything based on an Instagram ad.” By contrast, 16% of Facebook users bought something based on an ad. Now that's different research in the same article but given that those numbers:
- Are self-reported, as opposed to behavioral, and,
- Include internet users who are not active social media users
It’s pretty safe to assume Facebook’s influence is much higher. My own research and research from friends of mine like Rob Petersen certainly suggest so.
Facebook’s Data Problem: Summary
So, yes, Facebook has a data problem. It’s likely miscounting people for any number of reasons, most of them likely innocent at least as far as Facebook’s behavior is concerned.
If you’re looking for reach, Facebook’s data problem is your problem too. They need to understand exactly how many unique individuals your advertising will reach. That’s going to be a problem for the company as they look to expand their video offering to compete with YouTube via Facebook Watch.
If on the other hand, you’re looking to connect with a richly targeted audience, don’t worry too much about the numbers Facebook reports. Instead, look at your own numbers. See whether Facebook is driving profitable traffic and conversions through your own channels. If they are, does it really matter whether how many people they’re reaching? In fact, their miscounting might actually make Facebook’s performance look worse than it is by artificially increasing your impressions, and undercounting clickthrough rates.
The point is, don’t believe the hype. I’ve mentioned many times content is King, customer experience is Queen, and data is the crown jewels. This is a perfect example of why you want to use your data to see if Facebook is producing the results you want. It’s not great that Facebook has a data problem. But that doesn’t mean you have to make it be your problem too.
Now looking at the clock on the wall we are out of time for this week. I do want to remind you that you can find the show notes for today's episode as well as an archive of all episodes by going to Tim Peter.com/podcast. Again that's Tim Peter.com/podcast. Just look for episode 202. And while you're there you can click on the subscribe link in any of the episodes you find there so you get us delivered to your favorite podcatcher every single week. You can also subscribe in iTunes or the Google Play Music store or Stitcher Radio or whatever your favorite podcatcher happens to be. Just do a search for Tim Peter Thinks, Tim Peter Thinks Out Loud, or plain old Thinks Out Loud, we should show up for any of those. And if you'd be willing to provide a rating on iTunes, the Google Play Store, or Stitcher Radio while you're there, I would really appreciate it. You can also contact me by going to Facebook.com/TimPeterAssociates, on Twitter using the Twitter handle @tcpeter, or via email by emailing podcast@TimPeter.com again that's podcast@TimPeter.com. With that I want to say thanks again for tuning in, I really do appreciate it. I hope you have a fantastic weekend and I will look forward to talking with you back here on Thinks Out Loud again next week. Until then take care everybody.