Most companies are certain they're at L3.
Almost none are.
We've heard the same line in roughly 200 diagnostics: "We're pretty mature on this stuff already." Then we look at the portal. Then we run the AEO citation snapshot. Then we map the workflow library. Almost every time, the company is at L2.
That's not a failure. L2 is where the work begins. The point of the Maturity Model isn't to grade you — it's to locate you, so the next move is informed instead of accidental.
Five levels. Pick one to see the pattern.
Tools talk. Data is centralized. The model isn't yours yet.
- Working HubSpot portal that mostly captures contacts and deals.
- A list of integrations that "kind of work."
- A reporting layer nobody fully trusts because the underlying data is dirty.
- AI experiments that don't compound because there's no real substrate underneath.
- Properties proliferated. Workflows accumulated without curation.
- The data model is the default shape of the platform — not the shape of your business.
- Custom objects you'd benefit from haven't been built.
- Breeze turned on but not configured beyond defaults.
- Portals set up two to four years ago by a previous agency
- Companies that grew past their original kernel structure but haven't re-modeled
- Workflows built to solve a specific problem and never retired or refactored
- Properties added by ten different people across two years, no naming convention
- Cross-system reporting that requires manual assembly even when supported
- 01Custom objects built for your primary work units (matters, projects, engagements, accounts).
- 02Properties curated, named, typed, governed. Naming conventions documented.
- 03Associations modeling the real relationships, not the defaults.
- 04Workflows audited and rebuilt. Each one accountable to a specific signal or action.
- 05Data hygiene restored. Dedup, fill-rate floors, validation rules.
- 06Reporting rebuilt to produce decisions, not dashboards.
- 07Integrations wired to be queryable, not just connected.
- 08Breeze configured deliberately (just the start of what's possible).
- 09Brand-as-prompt library installed inside the kernel.
- 10Voice consistency tuned across at least the top three channels.
Three rules of thumb.
Take the assessment. Or skip the form and talk to a human.
10–15 questions. Five minutes. Returns your current level, the size of the gap, the recommended starting layer, and a downloadable report you can share with leadership.
A 90-minute, denser version of the same exercise, with two senior people who've located 200+ companies on the Model.
Map your company's intelligence in 90 minutes.
No slides. Just the map. We'll whiteboard which level you're on across each layer, and which moves get you to the next one.
Yours to keep — signed or not.

