CI Web Group is not a case study I downloaded. I founded it in 2006, became a Daikin preferred partner around 2008, rebuilt it from 320 people to 38, and today we do over ten million in revenue on an AI infrastructure and intelligence layer. These are the lessons that survived — not the ones that sounded good in a deck.
If you are building in trades, home services, or any relationship business being rewritten by AI search, these lessons transfer — not as copy-paste tactics, as pressure tests for your own shop.
Lesson one: specialize before you scale
Generic agency work is a treadmill. Trades marketing — HVAC, home services, the operators who keep communities comfortable — gave us standards, language, and partnerships that generic shops never earn. Specialization is how you become legible to customers and to AI search alike.
Lesson two: the agency model caps you
Selling hours and deliverables keeps you on a hamster wheel. Building outcomes and engines changes economics. I wrote the breakup letter in The Agency Model Is Broken and the build side in Revenue Engines, Not Websites.
Lesson three: systems and soul or neither
The thesis that runs the company — and the book — is Systems and Soul. Technology that scales care. Values that decide what technology is for. Client-first before autobiography — The Client-First Inversion.
Lesson four: people jobs changed, not just tool jobs
The restructure was the SME Inversion at company scale — experts to engineers to stewards. Three Hands is how we talk about it now. Chris built the layer; leaders govern it; I stopped being the answer key for every exception.
Lesson five: entity brand is infrastructure now
Your brand is not only how you look — it is what AI agents can verify about you. Inconsistent phone numbers, stale service pages, and reviews that do not match reality are operational failures with search consequences. We learned to treat entity truth like plumbing: invisible when it works, catastrophic when it does not.
Lesson six: speak honestly or lose the room
Keynoting for trades kept me honest. Owners smell consultant perfume. They respond to operator truth — including the cost of change. That is why I publish predictions with dates and write essays like From 320 to 38 instead of hiding behind "transformation" buzzwords.
What almost broke us
- Headcount as vanity metric — see How I Measure Business Success.
- Hero culture — Systems, Not Heroes.
- Treating AI as a tab, not architecture — AI is not ChatGPT.
- Slow CEO evolution — Founder vs CEO.
Proof in the market
Women in HVACR and Women We Admire invite me back because I speak from the P&L, not from trend reports. Keynotes for trades work when the speaker has been in the truck line of sight — partnership standards from Daikin years, entity files agents can read, predictions with dates on them.
The long-form version is Hands Up — a letter to my grandson Dallas and the industry. Chapter seven is the restructure. Chapter ten is optimize for both. Start wherever your company is stuck.
What I would tell 2006 Jennifer
Specialize sooner. Build systems before headcount. Treat entity truth as infrastructure before agents forced the issue. Evolve from founder to CEO before the market forced a restructure. Write the values down plain — Dallas deserves an industry that already learned those lessons.
Action
Pick one lesson above. Name the behavior in your company that contradicts it. Fix that before you buy another tool. Start with chapter seven if you are mid-restructure — it is the most honest operator account I have published. If you want the operator map for your shop, work with me. If your peer group needs the honest version of this story, book a keynote.