The challenge
The scope was enterprise-sized: 37 products, 19+ web properties, and 20 product teams, each with its own priorities. The headcount was not. Adding people one channel at a time would never catch up. The answer had to be a model that let a small team do far more, and a way to prove it was working.
The insight
Design the operating model first, then the team. When rules, context, data, workflows, and schedules are each defined once, people spend their time on judgment, not on gathering data and rebuilding reports.
"Scale never goes to the human. Judgment never goes to the machine."
What I built
A small team with clear ownership
I grew the function from one person to three, and own the budget and agency strategy, measured against pipeline ROI, not activity.
Standards that scale across 20 product teams
One naming and messaging standard, a URL-classification standard, and a shared data taxonomy, so every team feeds the website, search engines, and AI engines the same story.
An AI-native operating model
- 11 scheduled workflows run on their own: a daily AI bot access check, Monday reports and a dashboard rebuild, a weekly plan, daily update, and weekly close, and monthly benchmarks.
- 6 AirOps workflows handle repeatable production and analysis: blog briefs, YouTube optimization, AI product perception, competitive analysis, market research, and content strategy.
- 8 purpose-built Claude skills cover the judgment work: strategy and redlines, writing, schema, technical diagnosis, claim accuracy, data analysis, and executive decks.
One scoreboard
Weekly reports refresh on their own. Monthly reviews with product owners, starting with MFT, MDSS, and ICAP, turn findings into the top three actions.
A reusable blueprint
I've since turned it into a business-agnostic operating system, five layers applied at every level from the growth function down to a single skill, that any growth team can run.
OS Ecosystem visuals under Ideas
Swipe sideways to inspect the full diagram →
The results
Most-cited brand in our category's AI answers, ahead of CrowdStrike and Palo Alto Networks.
First-touch pipeline sourced by organic and AI search in H1 2026, including the company's first AI-search-attributed closed-won deal.
AI search conversion versus the site average, the top-converting channel.
Team growth while scope stayed at 37 products and 19+ web properties.
What carries forward
A lean team can operate at enterprise scale when the system does the routine work and people own the decisions. That's the model I'd build in the next role: operating model first, then hire into it.