I’m All-In on a Hyper-Personalization Identity Layer
The Future of the Internet is more private, more 1:1 conversations between brands and consumers, and data brokers and third-party identity…
The Future of the Internet is more private, more 1:1 conversations between brands and consumers, and data brokers and third-party identity list vendors left out in the cold.
The internet is very broken right now. Has your CEO ever asked you to use AI to anticipate customer needs? Has someone on your product or marketing team called out the difficulty of activating the data in your CDP to create experiences that help convert customers?
Right now, everyone is running around data warehouses, CDPs, and stringing together first-party and third-party data from across the internet in the hopes that somehow this actually delivers personalized experiences to consumers.
We spend billions of dollars figuring out who is a Spanish-speaking dog owner living within 30 miles of Philadelphia, Pennsylvania. Today, twenty person teams of $200,000-a-year+ salaried data scientists crunch first-party, second-party, and even zero-party data in Snowflake, Amazon Web Services, off Oracle, or even in layers and layers of Excel spreadsheets, in order to figure out whether Spanish-speaking dog owners living within 30 miles of Philadelphia, Pennsylvania have better return on ad spend on Facebook or Google.
What about privacy? Data collection challenges? GDPR? Do you even know what GDPR stands for, or do you just know the acronym? You better hire some lawyers at $500 an hour to help guide the data engineer who is going to build the data clean room so that the data scientists can build audiences based on events like clicks and web activity tied to third-party enrichment. Then, and only then, can you effectively target all of the Spanish-speaking dog owners living within 30 miles of Philadelphia, Pennsylvania.
Oh, and you have five competitors all doing the same thing, trying to find all of the Spanish-speaking dog owners living within 30 miles of Philadelphia, Pennsylvania.
Problem #1: Identity must be contextual
Most people are context switchers. For several hours a day they are commuters, then office workers, then commuters again, then busy moms or dads running kids to soccer practice, then Netflix watchers in the evening.
In one day, someone may switch from Gmail to booking dinner reservations on Resy to using Netflix, then switching to Hulu, then asking Google to play a favorite Spotify playlist.
The consumer state is always switching.
Right now, the identity profile of who someone is and what she likes, what she cares about, and what she buys is all held by data brokers, as they have the full profile on her.
“Moms who drink wine” or “Likely Democratic Voter” or “Single Asian Female” are the best these data brokers can come up with at scale. They are selling audiences with low context.
If there is a new bagel shop opening in town and the new bagel shop wants to target consumers, right now the best they can do is hope that “Moms who drink wine” and some data they bought about a competitor’s bagel purchase history is enough to target people.
Problem #2: Most of the data out there is bad anyway
Data brokers and identity vendors are playing a race to the bottom, and it’s only getting worse as the cookie goes away.
The cookie has been the best way in the AdTech era to string together behavior and activity across the internet, to tie purchases with events across a consumer’s internet experience. Apple has already killed most of this, and Google through the Chrome browser is next.
Match rates are down across the market. Several years ago, a brand could enrich their lists of customers and see 50%-75% of their customers match. Today, in some cases, that number is down to 20%-30%.
Data brokers and identity vendors are resting on their laurels, charging brand customers more money for increasingly worse enrichment. In fact, in all my scans and research about this space, many of these vendors are simply overselling bad data to fill out lists.
As a recent example, I took the Epsilon Data Challenge. Epsilon is one of the largest brokers out there.
- Go to https://legal.epsilon.com/dsr/ on Google Chrome if you have your privacy settings set to default.
- Click your country, click “Access my Personal Information”
- Submit request
My results are ridiculous.
When my partner, who is Ashkenazi Jewish, takes the test it believes she is an African-American woman.
Also, I am listed as someone who buys Adult Incontinence products. I cannot for the life of me remember when I have ever bought Adult Incontinence products, but perhaps it’s related to the fact that I am also listed as a frequent buyer of Snack Rolls, Asian Marinades, and Gala Apples. I would think that if someone is purchasing Snack Rolls, Asian Marinades, and Gala Apples together and consuming them for most meals, perhaps they would indeed wish to be advertised to about Adult Incontinence products, because Snack Rolls, Asian Marinades, and Gala Apples together seems like a recipe for disaster.
Vendors are simply monetizing increasingly poor lists when they sell data about consumers. The wool is getting pulled over the eyes of brands and all the surplus value goes right back to the vendors supplying these increasingly bad enrichment lists.
Brands lose, consumers lose, data brokers and identity vendors sit back and smoke cigars.
Problem #3: Implicit consent is not only creepy, but dying
For at least the last decade, implicit consent has driven a rudimentary form of personalization on the internet. It’s powered advertising and first party personalization like search and site optimizations for relevance.
As a simple example of implicit consent, when I access a website, pieces of me, a bunch of information from my browser is visible to the website. These things include their IP address, screen size, browser version, and cookie identifiers websites or identity resolution providers may have placed on my browser.
With increasing regulations and the end of the cookie, this is all over. Many forward-thinking brands have already moved to guiding customers toward a mandatory or “highly encouraged” login, and this is a first step toward explicit consent.
So what’s the solution? How ‘should’ the internet work?
That’s simple: an identity layer that sits between brands and consumers, with the application layer separate from back-end data and governance. Headless, hyper-personalization with language models generating personalized text, delivered on a consumer’s private and permissioned context.
Headless personalization is a data interoperability model that allows you to use any language model to generate personalized text with a consumer’s private permissioned context. It works by separating the application layer from the back-end data and governance. This allows customers to share (and un-share) their data to applications they love, while giving businesses the flexibility to transform and infer from it on demand. Apps permissioned to perform this headless personalization use inferences from connected consumer context and don’t directly handle the underlying data that powers it. Since consumers have more control over how and what they share and for what purpose, data sharing becomes chill.
Opt-in and interact with the brands you love. Don’t opt-in to brands you don’t love.
Zooming out the lens, this has major economic impacts.
- Retention, and especially Loyalty & Rewards, becomes king. Brands will be incentivized to deliver the best possible outcomes and experiences for their existing customers.
- Acquisition becomes a race-to-the-top instead of a race-to-the-bottom. Spam is reduced as brands no longer can afford to throw poor advertising based on shoddy enrichment data over the fence.
- Consumers will become more loyal to brands that deliver custom, hyper-personalized experiences and less loyal to brands that don’t deliver.
- AdTech as we know it collapses. There’s simply no more room for constantly reselling lists over and over again. The power is put back in the hands of the consumer.
I’m very excited about this new, potential future, and among new players in the space Crosshatch is an emerging player in the space building this more chill internet.
I’m speaking on this topic at Crosshatch’s Identity Con March 8th, at noon Eastern time. (You can sign up, for free.)
