For digital publishers: what owning your identity layer actually requires, and why surrendering your resolution layer puts you at risk.
If you run monetization or tech strategy for a publishing business, you've spent the last few years building a first-party data strategy. You collect logged-in emails, deploy first-party IDs, and integrate identity modules across your properties.
You think you own an identity asset. You don't. You built a supply line for someone else, and you're paying for the privilege.
The Illusion of "Proprietary"
There’s a fundamental difference between signal generation and signal resolution.
Generating a first-party signal — a logged-in user, a newsletter click, on-page intent — happens on your domain. Resolution, mapping that signal to a persistent ID that buyers can bid on, is typically where your control ends.
The distinction that matters isn't whether you rely on third-party infrastructure. Building your own stack from scratch isn't practical for anyone. What matters is how resolution happens and who controls the graph. Does the mapping stay governed by you (you set the terms for where it goes and what it's used for), or does it get absorbed into a graph the vendor owns, controls, and sells from?
Pooling and Reissuance: A Familiar Story
There's a precedent for what happens when the thing you originate gets pooled into someone else's inventory and sold on without you.
Think about how mortgage-backed securities work. A loan starts as a relationship between a borrower and a lender. The paperwork usually permits the lender to sell that loan. Most borrowers never read that far, and few think about it until the loan is three owners removed from the bank that made it. Once it's pooled with thousands of others and reissued, a new owner holds the asset and collects on it. The originating lender made the loan, took a fee, and dropped out. Every dollar the borrower pays after that goes to whoever holds the security, never back to the source.
Identity graphs run a version of the same play, and this part isn't hidden. It's a product line.
Vendors that build centralized identity graphs openly sell audience extension — your matched, deterministic signal becomes the seed for a lookalike audience that reaches new users who were never on your site, sold through the vendor's marketplace, on the vendor's terms. More than one major identity graph provider markets this as a named product line: match your first-party base, model a lookalike audience from it, and extend that audience across inventory you don't own and never see. This is documented product marketing, not a rumor.
Here's the part that gets lost in the fine print: you technically agreed to this. It's in the terms of service. But "agreed to" and "understood" aren't the same thing, and most publishers signed without mapping what resolution and extension actually transfer once your signal enters someone else's graph.
When your resolved signal is absorbed into a graph you don't govern, three things follow:
You're cut out of the deal on your own audience. Once infrastructure you don't govern maps your user, that audience can be reached without your inventory. You should be the one authorizing and monetizing the extension of your audience, even if the deal you choose is trading it for a fee, rather than discovering after the fact that it happened.
You hand over your training data advantage. Resolved inside a graph you don't control, whose models are learning from your audience? Your signal can train systems that price, target, and predict your audience, and those models keep improving on your data long after the impression clears. You can't see which buyers benefit, whose systems get smarter, or on what terms. In some cases the data never shares in the value of the media buy it helped win.
Your yield stays hostage. When an identity graph updates its logic, re-tiers how it prices your audience to buyers, or changes its integrations, your match rates fluctuate overnight. You’re left optimizing yield against changes you didn't approve and can't audit.
The alternative is to keep the resolution and the output under your control. When you own resolution, you decide what happens to your signal and capture the value when it's monetized. The vendor operates the rail. You govern what travels on it and where it goes.
True Ownership Requires Governed Resolution
Owning your identity layer isn't a legal distinction or a contract clause. It's a governance test:
Who controls what your signal resolves into, who authorizes where it travels after that, and does the output come back to you as an asset you hold, or disappear into a graph the vendor owns and sells from?
If you can't answer that, your signal is being pooled.
When you control resolution on those terms:
You control extension. You decide who reaches your audience and capture the value, across any currency a buyer wants to transact in. Extension happens because you authorized it and on terms you monetize, not as a byproduct a vendor captured invisibly.
Your signal compounds. As automated and agentic buying tools take over, your first-party signal becomes irreplaceable. The insights compound in your models, rather than in the models owned by centralized infrastructure.
Your yield stays protected. Buyers come to your inventory to reach your audience, preserving your pricing power.
Your audience signal is the most valuable asset you own. When you activate a currency through GrowthCode, you only transmit the signals you've authorized under your own agreement with that currency. Buyers transact solely on what they already know about that user, matched against your owned graph, not a shared one. If you want a look at how GrowthCode keeps that signal, and the value it creates, on your side of the ledger, that's worth twenty minutes. Request a meeting.
Jonathon Shaevitz is Co-founder and Head of Product at GrowthCode. GrowthCode provides identity and data infrastructure as-a-service for the digital advertising ecosystem.