contexture · live
visit· session views· total views· time on site· avg views/visit· days since last·
sample

behavioral profile

visitor type
·
first page of session
·
avg time on site
·
engagement
·

query yourself · CX.io()

CX.io(
//==> run a query to see your own context

about

I have been building this layer for twenty years. I just kept calling it other things.

Contexture is not a side project. It is the distilled output of a martech career spent solving one problem at the highest level: the absence of an owned, first-party context layer underneath everyone else's tools.

01 · the problem found me early

In 2004 I was the only JavaScript resource at Sprint.com, so all of analytics and martech landed on me. To keep nine tag vendors and a full SiteCatalyst install running across thirty-two properties, I built the world's first tag manager, years before you could buy one.

02 · I kept building pieces of the layer

At Sprint I stood up the first real data layer and invented ScrAPI, scraping a page's data as a pseudo-API to assemble personalized components live. Partnered with our head of optimization, Derek Gilbert, on the testing strategy, we lifted the flagship conversion rate by 408 percent. That work made me Adobe's first Technical Consultant hire, where I won the Founders Award for ScrAPI and built persistent-profile solutions for enterprise customers including Nordstrom, BMW, Verizon, and REI.

03 · then I hit the exact problem Contexture solves

Symantec hired me away to fix a stack their IT team said would take nine months. I shipped Adobe Target to production in three days. The hard part was the Norton renewal event: deliver a personalized offer on the first page of a session, to a visitor who had not identified themselves, or lose the moment. No tool could do it; analytics arrives too late. So I calculated the behavioral context myself, in profile scripts, for anonymous first-page traffic. That is the exact gap Contexture fills.

04 · I named it, then rebuilt it clean

At Ensighten, in the CTO's R&D team, I built an early version called Ensighten Context, where I first formed the Context Layer concept, and became a named co-inventor on US Patent 9,553,918 for anonymous, persona-based calculation. But it carried a heavy dependency. When I left, I rebuilt the whole thing from zero: no dependencies, every shorthand from a decade of front-end work, a tiny library that augments your vendors instead of replacing them.

05 · why I'm sharing it now

The plan was a startup; COVID ended that, and I moved my focus to AI architecture. So instead of a company, Contexture is the culmination of the career, packaged for the practitioners who can still use it. Context has never mattered more. AI can build almost anything, but not this, not from scratch, and the consultants doing real CRO and ABM still need an owned context layer underneath the tools.

the rooms I worked in. As Adobe's technical consultant I built solutions for Ally, Autodesk, BMW, Carfax, Experian, Macy's, McAfee, Nordstrom, Orbitz, REI, Stamps.com, Verizon, Virgin Mobile, and Symantec. At Ensighten I worked alongside United, Fidelity, American Express, T-Mobile, Microsoft, Dell, Staples, The Home Depot, and Safeway through the customer advisory board and beta program.

who it's for

You own the dataLayer and the tag manager. You read JavaScript fluently. You have felt the ceiling of a profile script and wanted real, portable, first-party context underneath your stack. If you live in the Measure Slack and at MeasureCamp, this was built for you.

three parts

One library, one book, one toolkit.

the library

The 6 KB SDK. Install it, query context with io(), syndicate it anywhere.

See the library →
the book

The field manual: the whole method, the recipes, the privacy model, and the full API reference.

Read the book →
the toolkit

Inspect and debug the context layer in the browser, on any site, then syndicate with confidence.

Open the toolkit →

This is the craft, distilled. The book teaches it; the library is it.