Context
A B2B/B2G technology company whose sales culture ran on relationships, handshakes, and text messages. Sales and leadership had no working concept of MQLs or SQLs. The company's proprietary platform had no data dictionary. The COO, whom I reported to, brought in outside RevOps consultants for a full CRM transformation in HubSpot. Everything about the client base is confidential, which is why this page has no screenshots and never will.
What Looked Broken
"We need the CRM implemented correctly." A tooling project with competent implementers. What could go wrong?
What I Noticed
The consultants were building from the official version of how the company operated — the COO's flowcharts, the documented handoffs. But I sat close enough to the actual work to see what the implementation process skipped. The people performing the work and the people building the system were never adequately in the same conversation. Requirements said one thing; the observable workflow said another.
What the Evidence Showed
Concretely: I kept intercepting conversations that were about to encode a misunderstanding, and sending clarifications to the RevOps strategist. "When she said X, she meant Y, because of Z — and the COO expects ABC." Nobody was lying. The operators described their work in the language of the official story, the consultants heard the official story confirmed, and reality stayed where it lives: in the work, invisible from above.
The Actual Constraint
The organization had no mechanism for reconciling how the process was supposed to work with how the work actually happened. The implementation was on track to encode aspiration rather than reality — competently. This is Coordination Friction in its purest form: every actor individually reasonable, the failure living entirely between them. And the fix, for a while, was me operating as unofficial infrastructure, which is a Shadow System with a job title.
My Hypothesis
If someone deliberately owned the gap — carried actual process knowledge to the implementers and shared definitions back to the operators — the system could encode the real business, and the definitions could outlive the project.
What I Architected
- The translation layer itself: structured clarification between operators, leadership, and consultants, so discrepancies were resolved before they became configuration.
- The company's first shared lead-lifecycle definitions (MQL, SQL, and the handoffs between them), taught patiently to a C-suite and sales team encountering the concepts for the first time.
- The beginnings of a data dictionary for a proprietary platform that had never had one, built in HubSpot so meaning lived in the system rather than in individual memories.
- Reporting dashboards giving marketing, sales, and leadership one set of KPIs to argue about instead of three.
What I Personally Built / Did
The clarification notes. The definitions and the teaching. The data dictionary work. Dashboard architecture in HubSpot. And a lot of unglamorous interception.
What Happened
The transformation shipped encoding the business that existed. Leadership gained its first shared pipeline vocabulary. And this project changed my career: it's where I first learned the term "RevOps," and where I realized the work I kept doing wasn't quite RevOps — it was something that needed a different name. I eventually called it Revenue Architecture.
What Became Possible
A sales conversation and a marketing report could refer to the same pipeline stage and mean the same thing. That sentence sounds small. It was not.
What I Learned
Implementation quality can't compensate for requirements that describe the wrong company. The connective tissue between operators and implementers is real work, it's nobody's job by default, and if you don't assign it, it gets performed invisibly by whoever can't stand watching the gap — until they leave.
Evidence / Artifacts
Nothing from this engagement can be shown, even redacted; the client space is genuinely sensitive. The diagrams on this page are recreated, generalized illustrations of the structure of the problem.
Related Work
The five-interpretations diagnostic from the same company: Documented ≠ True. What this work grew into: I Kept Seeing the Same Problems. So I Built a Diagnostic.