visitor type
·
the concept
The context layer, made visible.
A calculated, first-party context layer that sits beneath the data and personalization categories, computed once and portable everywhere. The idea this whole project is named for.
context·relevance·engagement
The borrowed-context problem
Every tool in the marketing stack calculates context, and every tool calculates it differently. "Engaged visitor" means one thing in your testing tool, another in analytics, a third in the CDP. The math gets rebuilt by hand in each platform, every engagement. And the result is captive: it lives inside the vendor that computed it, so switching tools leaves your context behind.
You end up renting your own behavioral data back from the platforms that happen to hold it. The context layer is the answer to that: calculate context once, on your side, in storage you own, and feed it to every tool consistently.
What a context layer is, and is not
It is not a tag management system. A TMS delivers tags; it does not build a persistent behavioral profile or calculate context. It is not a CDP. A CDP needs identity and stitches known records; the context layer works on the anonymous, unidentified traffic that never logs in, the majority of every site. It is not a testing or personalization tool. Those act on context; they do not own a portable, first-party definition of it.
A context layer is the thin, owned tier underneath all of them: persistent, anonymous, first-party, vendor-neutral. It is a first-party, headless CDP that fills the gaps, onsite, in the middle of the funnel.
Context, relevance, engagement
Context is what you know about a visitor's behavior. Relevance is the decision that context enables. Engagement is the result when relevance lands. The model runs in that order, and each beat depends on owning the one before it. Borrowed context breaks the chain at the first link.
Why now
Third-party cookies are collapsing, and with them the borrowed-identity model that personalization quietly depended on. First-party, anonymous, consent-gated context is what remains, and it is what this layer is built to own. The shift is not a threat to the approach; it is the reason for it.
The ownership and portability thesis
If you own the context, you can move it. Switch your testing tool and the engagement score comes with you. Add a new channel and it inherits the same affinities. The context layer makes your behavioral data an asset you carry, not a feature you rent.
source: the field manual, adapted to web reading.
principles
What you already believe, if you've felt the problem.
-
01
Own your context
Most of the data deciding what a visitor sees isn't yours. Reclaim that layer.
-
02
Compute once, syndicate everywhere
One definition of "engaged," calculated on your side, fed to every tool.
-
03
Context should outlive the tool
If switching vendors loses your behavioral data, you never owned it.
-
04
Fill the gap the vendors leave
CRMs and CDPs need identity, and third-party cookies are going away. The anonymous majority of your traffic falls through both. First-party context catches it.
-
05
Let context choose the subtext
Who they are and what they keep returning to tells you which message is relevant, and relevance is what earns engagement. Context, then relevance, then engagement.
"Context is the customer's side of the conversation. Subtext is the marketer's side."
-
06
Small enough to run first
6 KB gzipped, zero dependencies, frozen on purpose. Code that runs before everything else has no business being heavy or shifting under you.
the journey
From borrowed to owned
borrowed context
Every tool computes its own definition. The same visitor is three different people.
rebuilt by hand
The same math gets reimplemented in every platform, every engagement, and stays stranded there.
owned and portable
One definition, calculated on your side, in first-party storage, syndicated everywhere.
the stack
Five layers, one instrument
Integrated Orchestration
syndicate to any tool you own
Contexture integrates with 3rd party vendor solutions to take tactical action in real time
Calculated Context
scores · affinities · recency
Contexture calculates additional profile and segment context as a Context Layer
Persistent Profile
anonymous first-party identity
Contexture provides session and visitor level data storage for persistent user profiling
Abstracted JSAPI
one io() surface over everything
Contexture ingests your website data and provides a consistent data query API
Embedded Data
dataLayer · params · meta · cookies
Default metadata, variables, parameters, cookies, content exposed in your webpage
profile
- whotraits
- whatbehaviors
- howaffinities
segment
- whymotives
- wheremilestones
- whenthresholds
collect at the base · calculate in the middle · syndicate from the top