An SMB's AI maturity is assessed across 5 dimensions — data, processes, team, strategy, tools — each scored from 1 to 5, rather than on the single question "do we use ChatGPT or not". Running this diagnostic before signing up for a partnership (399 or 699 € excl. VAT/mo) avoids losing months piecing together what an hour of honest self-assessment reveals, and points you to the right tier.
Every month, we receive messages that all sound alike: "I want to start an AI Partnership with you. Essential or Premium, I'm not sure yet. What do you recommend?". The question seems simple. In reality, you can't answer it honestly without knowing where the company stands. Not where the leader stands in their understanding of AI — where the SMB itself stands, viewed as a system.
That's exactly what the AI Maturity Score measures. And it's the exercise we recommend every business leader do before signing up for a monthly AI Partnership — with us or anyone else. Not to add an extra step, but to make sure the money spent (399 or 699 €/mo excl. VAT at Codito) tackles the real issues, in the right order, at the right pace.
This article walks through the method we use internally: the 5 dimensions to assess, the concrete criteria at each level, the 3 typical profiles that almost always emerge, and how the Codito AI Partnership adapts to the score you get. By the end, you'll know where you stand — and which format fits you.
Why an AI Maturity Score, and not just "an audit"
The Codito AI Audit remains the most comprehensive tool for framing an AI initiative in an SMB: 2 to 6 weeks, 2 to 15 people interviewed, friction mapping, a costed roadmap, a 60- to 80-page deliverable. But not everyone can — or wants to — start there. Some leaders have already had an audit done elsewhere. Others have a tighter initial budget and want to jump straight into the AI Partnership. Still others want a quick self-assessment before deciding anything at all.
For all of those cases, the AI Maturity Score is an intermediate compass. It doesn't replace a full audit. It doesn't provide a costed roadmap. But it answers a simple question: "Across the 5 dimensions that truly matter for making AI work in an SMB, where do I stand today?". And that answer shapes everything else — including the Essential vs Premium decision within the AI Partnership.
The Score's other advantage is that it's dynamic. Once calculated, it becomes a barometer you track month after month. For our AI Partnership clients, the score is recalculated in every monthly report: you can objectively see which dimension is progressing, which is stalling, and where to concentrate the next effort. That's what keeps the AI Partnership from drifting into firefighting mode and turning back into a string of directionless meetings.
The 5 dimensions of an SMB's AI maturity
Across the hundreds of interviews conducted since Codito launched, we've seen that an SMB's real AI maturity plays out across 5 distinct dimensions. Distinct because they rarely advance at the same pace: one company can be very mature on tools (everyone has a ChatGPT Team license) and a disaster on data (nothing is structured, information travels by email). Another can be solid on strategy (the leader has a clear AI vision) and stuck on team (nobody is on board).
The trap is reducing everything to a single number. The right reflex is to score each dimension separately, on a scale of 1 to 5, then look at the profile. Here are the 5 dimensions.
Dimension 1 — Data
This is the foundation. Without accessible, structured data of at least minimal quality, no AI project holds up. The question to ask yourself: "If I wanted to feed an AI agent with the history of my clients, quotes, contracts, or job sites, could I do it in a few hours, or does it first require a three-month structuring effort?"
Criteria to assess:
- Accessibility: is your business data (clients, quotes, invoices, projects) centralized in a CRM/ERP, or scattered across Excel, Drive, individual email inboxes, paper?
- Structure: are key fields filled in consistently, or does each salesperson name things their own way?
- Quality: when was the last serious cleanup of your database? How many duplicates, obsolete records, empty fields?
- Sovereignty: do you know where your data lives (cloud, on-premise, third-party provider), who has access to it, and what you can or cannot do with it using an external AI model?
Dimension 2 — Processes
AI slots into an existing process. If that process is neither documented nor stable, automating part of it amounts to automating chaos — which just makes it faster. The question to ask yourself: "Are my 3 most critical business processes (quoting, invoicing, client follow-up, onboarding, etc.) described somewhere clearly enough for a new hire to execute them without asking a veteran?"
- Documentation: do you have written procedures, or is everything in the heads of a few key people?
- Repeatability: do your processes run roughly the same way from one case to the next, or is each one handled "by feel"?
- Measurement: do you track simple indicators (response time, conversion rate, acquisition cost), or do you steer by gut feeling?
- Bottleneck identification: could you name the 3 precise places where your teams lose time every week?
Dimension 3 — Team
AI in an SMB isn't an IT matter. It's a human matter. The best friction map and the best agent in the world are useless if the team doesn't take ownership of them. The question to ask yourself: "How many people in my company, besides me, actively use an AI tool today — and how many would be ready to join a structured effort?"
- Current skills: who on the team knows how to write a useful prompt, read the output with a critical eye, and integrate it into their work?
- Openness: does the team see AI as a threat, a fad, or an opportunity to offload repetitive tasks?
- Available time: do you have the bandwidth (in-house or via an external partner) to lead the effort, or is everyone already at 110% on day-to-day operations?
- Designated champion: is there someone in the company who owns the AI topic, or does nobody truly carry it?
Dimension 4 — Strategy
This is the dimension leaders most often overestimate. "Of course I have an AI strategy". When you dig in, the strategy amounts to two vague statements ("We want to be at the cutting edge", "AI will change everything") with no concrete direction. The question to ask yourself: "If one of my managers came to me tomorrow morning with a 15k€ AI project, could I answer yes or no in 5 minutes, with a clear criterion?"
- Vision: do you have a documented intuition of what AI can or cannot change in your industry, over the next 12-24 months?
- Business objectives: have you identified 2-3 concrete outcomes you expect from AI (time savings, quality, ability to scale) rather than a vague ambition?
- Expected ROI: do you have a sense of what one euro invested in AI should return, and over what timeframe?
- Prioritization: could you rank your potential AI initiatives by impact, or is everything on the same level?
Dimension 5 — Tools
Deliberately last — because it's what leaders look at first, when it's actually the least critical dimension. An SMB can have 0 ChatGPT licenses, score high on the other 4 dimensions, and get started with AI very quickly. The reverse is rarely true.
- Current stack: which AI tools are already running in your company (ChatGPT Team, Copilot, Claude, business tools with embedded AI), and at what level of usage?
- Integrability: do your business tools expose APIs or connectors that would let an AI agent read from and write to them?
- Technical debt: do you have legacy tools (a 2008 ERP, an Access database maintained by a single person) that would block part of your AI initiatives?
- Switching cost: if you had to swap out a tool tomorrow to gain AI compatibility, could you do it, or would it be a 12-month undertaking?
The scoring grid: how to rate yourself honestly
For each dimension, give yourself a score from 1 to 5 using this grid. The trap is over-rating yourself out of optimism or under-rating yourself out of excessive humility. The rule we apply in audits: at any level, if you hesitate, take the level below.
- 1 — Nonexistent. The dimension isn't covered at all. For data, that means: everything lives in emails and personal folders. For strategy: no AI thinking has been done. For team: nobody uses an AI tool.
- 2 — Embryonic. A few elements exist, but in a disorganized way. For data: a CRM exists but is barely filled in. For processes: a documented procedure here and there, the rest passed along verbally. For tools: 2-3 AI licenses handed out with no usage plan.
- 3 — Functional. The basics are in place and hold up day to day, but without deep optimization. For data: a properly maintained CRM, an ERP used seriously. For team: 30-40% of employees use AI, without structured training.
- 4 — Structured. The dimension is actively managed, with indicators and a direction. For strategy: a documented 12-month AI roadmap. For processes: the 5 critical processes are described, measured, and a designated owner keeps them evolving.
- 5 — Advanced. The dimension is a competitive asset. For data: everything is structured, accessible via API, with quality tracked. For team: a genuine AI culture, ongoing training, several people capable of self-installing simple agents.
You end up with 5 scores, and therefore a profile. Not an average — a profile. It's this profile that guides the decision.
3 typical profiles — and the AI Partnership that fits each
Across the assessments we've run in audits or pre-audits, three profiles emerge almost every time. For each one, here is the typical score, what it reveals, and the AI Partnership format that fits.
Profile 1 — Low score (most dimensions at 1 or 2)
Symptoms: the leader reads articles about AI, sees competitors starting to talk about it, but nothing inside the company has moved. Data scattered between Drive and email inboxes, undocumented processes, a team that has never received AI training, no written strategy, ChatGPT used by the leader themselves when they find the time.
What it means: there is significant groundwork to do before serious AI agents can be installed. The AI Partnership alone can get things started, but it will spend its first 3-6 months piecing together what an audit would have laid out in 3 weeks.
The right answer: an AI Audit first (Express or Focus Pack depending on company size), then a transition to the Essential AI Partnership at 399 €/mo excl. VAT to execute the roadmap over time. The audit + 6 months of partnership combination produces far more value than a blind partnership over the same period.
Profile 2 — Mixed mid-range score (between 2 and 3 on most dimensions)
Symptoms: the SMB has already started moving. A properly maintained CRM, a few people use ChatGPT daily, the leader has a general direction ("this year we automate quoting and client follow-up") but no detailed plan. Processes run, data is accessible, the team is open. This is the most common profile.
What it means: there is enough substance for the AI Partnership to produce value quickly, without necessarily going through a full audit first. The risk isn't the framing — it's losing the thread between two quarters and the roadmap ending up back in the drawer.
The right answer: the Premium AI Partnership at 699 €/mo excl. VAT straight away, because the main value lies in getting the team on board (monthly training calibrated to their tools) and installing 1 to 2 simple AI agents they'll then know how to keep alive. Leadership coaching remains useful, but it's the team that unlocks the business lever. For this profile, the combination of coaching + training + execution built into the Premium AI Partnership is exactly what's needed.
Profile 3 — High but uneven score (3-4 on strategy and tools, 2 on data or processes)
Symptoms: the leader has a structured AI vision, several tools are in place, the team already uses AI on a few use cases. But when you dig in, the client database carries 8 years of technical debt, or the quoting processes aren't documented at all. The imbalance shows: the "AI surface" looks great, the foundation is fragile.
What it means: the next AI initiatives will run into the weak dimension. There's no point adding a fourth agent if the data it has to work with isn't reliable. The AI Partnership must prioritize strengthening the weak dimension, not stack AI projects on a foundation that's cracking.
The right answer: the Essential AI Partnership at 399 €/mo excl. VAT in strategic-coaching mode for the leader, to prioritize strengthening the foundation, possibly with a targeted Custom Development project on the weak dimension (CRM overhaul, product database structuring, quoting process redesign). Don't start Premium until the weak dimension has climbed back to 3.
Why assessing maturity changes the sales conversation
Let's be honest: without this assessment, the sales conversation around the AI Partnership quickly turns into a pitch — "Essential or Premium, look at what's included, pick the tier that suits you". With the assessment, the conversation becomes a diagnosis — "Here's what I see in your situation. Here's the weak dimension I see. Here's the format that tackles that dimension first, not the other one".
It's the difference between selling an offer and guiding someone toward the right decision. And that's why we build the assessment into the start of every scoping meeting — whether it leads to an AI Partnership with us, to an Audit, or even to another agency if we're not the right partner for that specific SMB.
To understand how the AI Audit, AI Partnership, and Custom Development fit together, we wrote a dedicated article comparing the AI Audit vs the AI Partnership. And to grasp how coaching and training fit into the monthly format, the details are in the article on the differences between AI training, coaching, and partnership.
How we actually do it at Codito
The assessment happens at several levels depending on your entry point:
- In a free scoping call (30 minutes): we walk through the 5 dimensions during the conversation, and you leave with a single-digit score per dimension and an offer recommendation. No formal deliverable — it's honest guidance.
- In an AI Audit: the AI Maturity Score is one of the central deliverables of the Focus, Essential, or Max Pack. It's calculated after 2 to 15 people interviewed in the field, based on factual evidence (not the leader's self-reporting), and presented in the final report.
- In an AI Partnership: the score is recalculated every month in the monthly report. You can objectively see which dimension has progressed, which has stalled, and where to concentrate the next effort. That's what keeps the AI Partnership from turning back into a string of directionless meetings.
No magic promises. The assessment doesn't work miracles. But it avoids the worst AI mistake an SMB can make: paying for the right format at the wrong time, or the wrong format while thinking you chose the right one. That's what costs the most — more than the tool itself.
In summary
An SMB's AI maturity doesn't come down to "we use ChatGPT or we don't". It's measured across 5 distinct dimensions — data, processes, team, strategy, tools — which rarely evolve at the same pace. Scoring each on a scale of 1 to 5 produces a profile, and it's this profile (not a single number) that guides the decision: Audit first, Essential AI Partnership straight away, or Premium AI Partnership with a focus on the weak dimension.
The reflex to avoid: choosing an AI Partnership format because it fits the budget, without having honestly looked at your maturity profile. The reflex to adopt: set aside 30 minutes to walk through the 5 dimensions, identify the weak one, and choose the format that tackles it first.
Want to assess your SMB's AI maturity?
Book 30 minutes with Emile: we'll walk through the 5 dimensions together, and you'll leave with an AI Maturity Score and the offer recommendation that matches your profile — Essential or Premium AI Partnership, or another format.