This is the unabridged version of an article that appeared in AdExchanger under the title ‘Telco Ad Tech Investments Follow A ‘Wheel Of Fortune’ Philosophy’
Yesterday’s news of the Tapad acquisition by the Norwegian mobile carrier Telenor is a complex play for the cross-device crown. Allison Schiff’s piece offered a sophisticated take on why these telecommunications companies (Telcos) are getting into the space. The general feeling in the industry is that we are on the cusp of a new era, one in which Telcos are realizing their potential by acting on the writing on the wall, one that’s got “churn to reduce, average revenue per user to juice and dumb pipes to monetize.”
It appears like every investor is running from digital advertising, but a fundamental truth is that digital is growing. Hell, spending in digital advertising will soon surpass TV. There are very few things you can count on, but one of them is that the role of data, the life-force behind marketing and advertising, is front and central to future success. Sure, the chatter may be about the VCs running for the hills, but everyone knows: there will be more winners. And who will the winners be? The companies with the most data: Google, Facebook, Criteo, Amazon, etc. are all good examples of early winners. These incumbents are able to create more revenue from each user than any traditional digital company ever has because of their data. There’s another truth: In a world of infinite inventory, and fragmented attention and device usage, the old way of marketing and advertising to devices is becoming less effective. It turns out, although it may not be obvious, the people data that makes Google, Facebook, Criteo, and Amazon powerful is the same people data needed to build an effective cross-channel and cross-device strategy. In 2016, there is no doubt: Data is king. Let me explain:
Data was obviously always important for Amazon, Google, Facebook, and Criteo to deliver their services.in this next era, where the average person is expected to have 4.3 devices by 2020, is new data necessary? Fortuitously, it turns out that the very same data that got the incumbents to where they are is also the key to building and wielding a cross-device and cross-channel identity (or no identity) graph. This concept of the graph is what most experts see as the future of the industry. Logins and log files, the building blocks of Identity and No Identity Graphs, is contained within the original data set of the incumbents. All you need to do is just add algorithms and business rules to the data, or, in the case of Verizon/Telenor, an acquisition.
So, the winners of the future, like the early success stories, will be skilled in combining data, acquiring it, and deploying it. And the reason why can be best explained by the classic game show Wheel of Fortune.
Deterministic vs. Probabilistic
As background, nearly every company in our industry can be put in two camps: A player in the open web (Rubicon, Criteo, Oracle, Mediamath, AppNexus etc.), or the closed web (walled garden operators like Facebook, Twitter, Verizon/AOL, Google, etc.). But what both camps struggle with is being sure of the identity of the customer in cross-device and cross-channel marketing efforts. Even Facebook, who has incredible logins across desktop and mobile, the basis for a deterministic graph, still has incomplete data.
Which brings me to probabilistic.
Can I buy a consonant?
A probabilistic identity match, as it’s known in the industry, is the way to identify a user without login data. In other words, a probabilistic match is an educated guess at the user behind the device. Probabilistic graphs give you scale, and can be used to fill in the gaps, the same way contestants would buy consonants in Wheel Of Fortune to get closer to the answer. But probabilistic is limiting: You can’t do measurement and attribution down to the person because guesses never walk into stores and buy things. Additionally, a shortcoming of probabilistic is that it prohibits you from doing things that require affirmative consent: like honoring a suppression or targeting a pre-approved audience.
Conversely, if you are a large retailer, and your goal is to re-engage with users who had demonstrated “live intent” (the industry term for intent data honed in real-time) for a product or service, relying solely on a deterministic match is a hindrance to scale.
Deterministic data is accurate up to the point you can trust the login is who you expect them to be. Deterministic data is accurate, but most companies don’t have it at scale. In the best scenarios, the same technology provider has both deterministic and probabilistic data and only deploys probabilistic when they don’t have deterministic data and when accuracy is not imperative – like guessing the phrase when you have 50% of the letters in Wheel of Fortune. There are times when you want hyper-accuracy (let’s say for a pre-approved offer) and there are times when you want reach (running a retargeting campaign that gets 10x the performance.) So, when Telcos make moves to have both, they have a tremendous advantage.
This is the biggest conversation in the industry right now. We saw it at Industry Preview 2016 in a panel where Omar Tawakol, SVP and GM of Oracle Cloud, discussed getting it right versus getting it right at scale. Why do Google and Facebook make so much more per user than everyone else? Because their data provides more resolution, allowing them to be infinitely more profitable than the average publisher. I myself fully believe in combining the probabilistic with the deterministic. I was an early angel investor in Tapad, am an early pioneer of people based marketing, and I have led LiveIntent to acquire a company with probabilistic capabilities to supplement our deterministic ones. I’ve put my money where my mouth is. If you think I’m wrong about this post, please let me know in the comments section.
I’ve begun to adopt the Wheel of Fortune game as the rubric when I explain why Telcos, among others, will fuel the next big wave of marketing innovation in our industry. Some companies may have bet their marketing strategy entirely on probabilistic models (consonants) and others may have bet entirely on deterministic (vowels), but either way, their strategy for guessing the answer will be lacking one or the other.
Telcos Have Done This Before
With Telenor acquiring Tapad and Verizon acquiring AOL, you might think this is the dawn of the Telco jumping into the advertising game. But these aren’t unchartered territories. This has been done before. One could say the first great Telco acquisition was in 2006, when Softbank (who have a joint venture with Yahoo! as Yahoo! Japan) bought Vodafone Japan. Yahoo! Japan has long had the reputation in the industry for being innovative. Whether they had a hand in the Vodafone acquisition and were years ahead of the west in building an identity graph is unknown…
Of course, in this I am truly an outside spectator, with no known knowledge of their motivations. But to me, it makes perfect sense, as Alison Schiff wrote about Telenor,
Singtel probably also had “churn to reduce, average revenue per user to juice and dumb pipes to monetize.”
In 2016, Hangman May Replace Wheel of Fortune
Wheel of Fortune, really, is just a less macabre version of Hangman, the children’s game where each missed guess results in another step closer for the condemned to be hanged. But maybe Hangman is the better comparison in 2016? Wheel of Fortune is fun, breezy, and led by the inane blatherings of Pat Sajak. Ad-Tech has had periods of being full of growth, of the same players that fail upwards, and of even the most poorly designed of innovations being acquired. But in a world of disappointing IPOs and desultory stock performances, the comparisons in 2016 may be more grim, and we’ve moved from Wheel of Fortune whimsy to Hangman realities..
In Hangman, there are three possible outcomes: you can sit on the sidelines and not guess, which guarantees you are no better than hanged. You can guess incorrectly until you are hanged, or you can take an educated guess and win. What we’re sure to discover soon is those Telcos who take action are likely to do very well as the market shifts to marketing to people because mobile has made this a necessity. Those that do nothing are destined to be hanged, slowly…
We’d All Like to Solve the Puzzle, Pat
The obsessives at Slate broke down the math of Wheel of Fortune and, unsurprisingly, the more letters you have before you guess, the better off you perform and the more money you make:
This is the same dynamic that will rule the future of our industry as we create solutions for the cross-device challenge. The more clues we have into the identity or anonymized identity of someone no matter what device they are on, the better off the cross-device marketing will be. In the examples of the Telcos, Verizon and Telenor had the vowels of deterministic data (the actual devices and their respective, anonymized IDs.) They recognized the biggest opportunity was to have the option to “dial-up” accuracy or scale to be able to best server marketers cross-device and channel, and they went out looking to buy some consonants. These consonants took the form of AOL and Tapad, respectively. Buying letters to get closer to solving the puzzle is the most obvious parallel to how cross-device is like Wheel of the Future, but it’s just the Tip _f th _ Ic_b_rg.