# The AI Maturity Model: Five Levels, and Where Most Marketing Teams Stall

> Where does your marketing team really sit on AI? The five levels, the six capability areas scored, and why most teams stall at Level 2.

Source: https://www.hiaitus.ai/posts/ai-maturity-model-five-levels

_Published 2026-08-06_

**Most marketing teams think they are ahead on AI. Few can prove it.**

Ask a marketing leader whether their team uses AI and the answer is almost always yes. Ask which parts of the work AI has actually reached, who owns it, and what it has changed in the numbers, and the answer gets vague. The AI Maturity Model exists to close that gap. It turns a feeling, we are doing AI stuff, into a position you can defend: we are Level 2 of 5, strongest in personal use, weakest in email and CRM.

## What is an AI maturity model?

An AI maturity model is a ladder describing how deeply AI is embedded in the way a team works, from occasional individual use through to agents running whole workflows. Ours has five levels, and every result is also scored across six capability areas, so you learn which part of the function is holding you where you are.

One thing it is not: a count of your licences. The gap between teams is rarely the tools. It is shared standards and measurement.

## What are the five levels of AI maturity?

Five levels, one word each, from Reactive to Autonomous.

- **Level 1, Reactive. **Ad hoc, no shared use. AI is used occasionally, by accident more than design. No expectation, no standard, no owner. Any value is incidental.
- **Level 2, Experimental. **Individuals experiment. A few people are ahead on their own initiative. Value arrives one person at a time, and it leaves when they do.
- **Level 3, Integrated. **Shared standards hold. AI is part of the standard process, with named owners. The team uses shared tools daily and there are agreed ways of working, so quality does not depend on who picked up the task.
- **Level 4, Optimised. **Measured and improving. The work is instrumented. Outputs are measured, compared and improved deliberately rather than assumed to be better.
- **Level 5, Autonomous. **Agents run the work. Agents run whole workflows and the team reviews the output. Humans set direction and hold the review gate; execution is largely delegated.

Each level is written so a team can recognise itself. Read them as sentences you would say out loud about your own team. The honest answer is usually obvious within seconds.

## Why do most marketing teams stall at Level 2?

Because Level 2 asks nothing of the organisation. One curious person with a subscription creates Level 2. Nobody has to agree on anything, write anything down, or change how the work moves.

Level 3 is the first level that costs something, and none of it is a tooling upgrade:

- AI becomes required, not optional. It is a step in the process, not a personal choice.
- Each use case has a named owner.
- The standard is written down, so the whole team follows the same one.

That is why two teams with identical tool stacks can sit two levels apart. Individual brilliance does not compound. A shared system does.

Most teams we speak with sit between Experimental and Integrated. We will publish benchmark figures once enough audits are in, not before.

## Which six capability areas does the audit score?

A single number hides the useful information, so every result is graded 1 to 5 across six areas.

- **Personal AI use. **How capably individuals use AI in their own work, independent of any team standard. Usually the strongest area.
- **Team adoption. **How widely and consistently AI is used across the team, versus by a handful of enthusiasts.
- **Strategy. **Whether there is a deliberate plan, with use cases, owners and intent, or just activity.
- **Workflow integration. **Whether AI is embedded in how the work gets done, or bolted on beside it.
- **Measurement. **Whether the impact of AI on output and results is measured at all.
- **Email and CRM. **AI use in lifecycle marketing specifically: segmentation, sequences, subject-line testing.

The pattern this exposes most often is high personal use and low measurement. Individuals are ahead; the organisation cannot prove any of it. That combination feels like progress and compounds into very little, because nothing is shared and nothing is instrumented.

Email and CRM is its own area for a reason. It is consistently the least touched part of the function, and consistently the fastest place to prove value, because the feedback loop is short and the numbers already exist.

## What does a result look like?

Here is a worked example. Treat it as an illustration of the format, not as a benchmark.

- Personal AI use: L3
- Team adoption: L2
- Strategy: L1
- Workflow integration: L2
- Measurement: L1
- Email and CRM: L1

**Headline: **Level 2, Experimental. **Strongest area: **Personal AI use. **Weakest area: **Email and CRM. **Overall: **1.7 of 5, an illustrative figure.

The written interpretation reads: you are ahead as an individual, but AI is not yet a shared system. Value arrives one person at a time, and email and CRM are largely untouched. The next step is to make one core process AI-led, with standards the whole team follows.

## What do you do with the result?

Every result ends in three actions, under the tagline we have used since the start. Pause. Think. Do. Narrow, then plan, then prove. For a Level 2 team they read like this.

**Pause. **Name one marketing process to standardise. Pick the workflow with the most repetition and the least risk, and make AI a required step, not an option.

**Think. **Write a one-page AI plan for the quarter. Three use cases, an owner for each, and how you will measure the output. Keep it light and shared.

**Do. **Start with email and CRM. It is your fastest place to prove value. Add AI to segmentation and subject-line testing, with a human review gate.

The order is deliberate. Narrow the focus before you plan it, and plan it before you try to prove it. The Do step always points at the weakest area with the fastest feedback loop, because the first win has to be provable.

## How long does the audit take?

11 questions, under 4 min. You see your level immediately, and no email is needed to see your result.

The instruction on every question is the same, and it is the only reason the result is worth anything: pick the statement closest to today, not the one you are aiming for.

If you want the full report, the six-area breakdown, the written plan and a shareable card, we ask for an email at that point, after the result has been delivered. One email with your report. We will not add you to a list without asking.

An Organisation AI Audit, covering operations, delivery, data and governance beyond marketing, is coming.

## Frequently asked questions

### What is the AI Maturity Model?

It is a five-level ladder describing how deeply AI is embedded in a marketing team's work: Reactive, Experimental, Integrated, Optimised, Autonomous. Every result is also scored 1 to 5 across six capability areas, so the level arrives with a specific reason attached to it.

### How do I know which AI maturity level my team is at?

Take the Marketing AI Audit. 11 questions, under 4 min, and your level is shown immediately with no email required. Each question offers five statements written in the first person, and you pick the one closest to today rather than the one you are aiming for.

### What is the difference between Level 2 and Level 3 AI maturity?

Level 2 is individual: a few people experiment on their own initiative. Level 3 is shared: AI is a required step in the standard process, each use case has a named owner, and the standard is written down. The jump is organisational, not technical.

### Why does the audit score email and CRM separately?

Because it is consistently the least touched part of the marketing function and the fastest place to prove value. Segmentation, sequences and subject-line testing already carry numbers, so the effect of adding AI shows up quickly and defensibly.

### Do you publish AI maturity benchmarks by industry?

Not yet. We will publish benchmark figures once enough audits are in, not before. Qualitatively, most teams we speak with sit between Experimental and Integrated.

## Ready to move up a level?

We help marketing teams make AI a shared system, email and CRM included. To start a conversation, email [stusheridan@hiaitus.ai](mailto:stusheridan@hiaitus.ai).

_Human Intelligence × AI._