Before any team builds anything with AI, there is a step almost everyone skips: actually reading where the function is right now. Not the aspirational version. Not the version that lives in the strategy deck. The version that exists when you walk through the team's week, hour by hour, and notice what is actually happening.
The cost of skipping the diagnostic is not subtle. It shows up four months later as a half-built workflow nobody adopted, a tool the team works around, a champion who burned out trying to drag the function somewhere it was not ready to go.
Reading first. Building second. Always.
Two axes before six
Before the six-axis diagnostic, run a faster read first. Score the team on two axes from 0 to 10.
Tooling. What can your team physically do with AI today? At zero, they have heard of ChatGPT. At three, they paste in context and use Custom GPTs. At five, they are building light automations in Zapier or Make. At seven, they can debug API calls and build self-updating dashboards. At nine, they are deploying small internal agents wired to your stack. At ten, they probably should not be in HR any more.
Strategy. What is the AI actually doing inside the function? At zero, nothing. At two, low-stakes time savers like rewording job ads. At five, AI use is encouraged, a champion has built playbooks, KPIs are starting to attach. At seven, AI is the default in some core areas: performance, onboarding, internal comms. At nine, your AI-enabled People workflows are influencing other departments. At ten, AI is how the function delivers.
Forget the number. Look at the gap. Tooling at six and strategy at two means the team can build but is building the wrong things. Strategy at five and tooling at two means leadership has bought the story but the team cannot ship. Both look like progress on a slide and feel like nothing on a Tuesday.
Once you know roughly where you are on the two axes, the six-axis diagnostic tells you what to actually do about it.
The six axes
For a People function, there are six axes that matter. None of them is about the model. All of them are about the team and the work.
1. Data hygiene
Where does the truth about your people live? In one HRIS, or three spreadsheets, or a Notion page maintained by one person who left? The answer determines what you can build.
A function with one clean source of truth can build retrieval-grounded workflows that actually work. A function with three sources cannot. Until the data is unified, AI will produce confidently wrong answers, which is worse than no answer at all. Fix the source before you fix the workflow.
2. Process clarity
For each major workflow, sourcing, screening, interviewing, onboarding, performance, comp review, can the team draw it on a whiteboard in five minutes? If yes, that workflow is buildable. If no, the workflow needs to get simple and mapped before AI touches it. Often the diagnostic conversation alone produces this clarity, and the building gets easier as a side effect.
3. Tool fragmentation
How many tools does the average workflow touch? Three is workable. Seven is not. The ceiling on what you can automate is set by how much glue you can build between systems, and the glue gets brittle fast above five integrations. A function trying to "do AI" while running thirteen disconnected tools should consolidate first.
4. Curiosity distribution
Who in the team has, on their own time, played with AI? Who has not? A function with three or four already-curious people across different sub-teams can stand up the champion model in a quarter. A function with one curious person and twelve sceptics needs a different approach, usually a ground-up enablement programme before any building starts.
Some people are ready to build now. Others need to see it work first. Both groups exist in every team. The diagnostic finds them.
5. Sponsor presence
Is there a named senior leader, CPO, Head of People, COO, who will publicly say "this is part of how we work now," and who will defend the time the champions take? Without that, every workflow build will be the first thing dropped when a quarter gets hot. With it, the work survives the first crisis.
The presence of a real sponsor is the single best predictor of whether anything gets built. Better than budget, better than tooling, better than capability.
6. Risk posture
How does the company think about AI risk? Is it a "we won't touch it until we know it's safe" culture, a "move fast and ask forgiveness" culture, or somewhere in between? The right governance model, and the speed at which you can build, is set by this, and pretending otherwise produces friction nobody anticipated.
A regulated company moves differently from a 60-person scaleup. Both can build. The shape of the build is different.
Reading the result
The diagnostic produces a six-line summary: a short, honest description of where the function actually stands on each axis, and what that implies about the order of the work.
A typical reading might look like:
Data: one HRIS, mostly clean. Process: TA and onboarding clear, performance murky. Tools: nine in active use, three obsolete. Curiosity: four people leaning in, two TA, one HRBP, one Ops. Sponsor: CPO engaged, Board curious. Risk: cautious culture, EU AI Act exposure.
From that, the build order writes itself. Start with TA workflows where the data is good and the process is clear. Use the four curious people as initial champions. Set up governance early because the risk posture demands it. Leave performance work for phase two, after the process gets cleaned up.
The diagnostic makes the strategy obvious. That's the whole job.
Capability maturity, by function
A useful side effect of the diagnostic is that it lets you compare People against the rest of the business. Most companies have wildly uneven AI capability across functions. Engineering at "adaptive," Marketing at "capable," People at "unacceptable." If you do not know where you are relative to your peers in the business, you will either over-promise or get out-flanked.
A simple frame, borrowed from capability maturity work in adjacent fields:
- Unacceptable. Refuses or ignores AI tooling.
- Capable. Uses AI for individual tasks: drafting, summarising, light analysis.
- Adaptive. Embeds AI into core workflows with human-in-the-loop checkpoints.
- Transformative. AI changes the operating model, not just the tasks.
You do not need every function at "transformative." You need to know where each one is, and to set the right next step for each, this quarter. For the cross-function vocabulary that connects this maturity ladder to the named industry models (Gartner, MIT, BCG), see AI maturity frameworks for G&A leaders.
The trap of skipping
Skipping the diagnostic feels like speed. It is the most expensive form of slow. Every workflow built without reading the function first carries a small bet, that the data is good enough, that the process is clear enough, that the team will adopt, and most of the bets lose.
Read first. The building is faster afterwards, every time. The first concrete build, in almost every case we have seen, is setting up the team's AI workspace so that subsequent workflows have somewhere to live. From there, the domain map tells you where to point the build effort, and the move from prompts to systems tells you what the work looks like once you start.
Grading the team was never the point. Knowing what to build next was.
What this connects to
Auto-recommended next reads in the People Ops cluster, ranked by shared concepts and headings:
- Designing the AI-native People team
- An AI enablement operating model for People leaders
- AI governance for People teams
- From prompts to systems
Common questions
- Why diagnose readiness before building anything?
- Because the cost doesn't disappear if you skip it, it just moves downstream and grows. Four months out it shows up as a rebuild, a tool nobody opens past week three, a champion who quietly stops pushing. Read first. It's cheaper every single time.
- What is the difference between tooling maturity and strategy maturity?
- Tooling maturity is what your team can physically do with AI: prompt well, build a Custom GPT, wire an automation, deploy a small agent. Strategy maturity is whether any of that is doing something that matters. Most teams are mismatched, and it rarely shows up on a dashboard. A team at six on tooling and two on strategy can build a beautifully slick Zapier chain that automates the wrong process end to end.
- What axes should we actually assess?
- Six, and none of them about the model. Data hygiene: one source of truth or three. Process clarity: can the team draw the workflow on a whiteboard in five minutes. Tool fragmentation: three tools per workflow or thirteen. Curiosity distribution: who has played with AI on their own time and who hasn't. Sponsor presence: a named leader who'll defend the champions' hours. Risk posture: cautious culture or move fast. Score all six honestly before you touch a single workflow. Skip one and the diagnostic just tells you a comfortable half-truth.
- Which axis matters most?
- Sponsor presence, and it isn't close. Everyone assumes it's budget or tooling. Watch what actually happens: a Head of People says out loud, in a leadership meeting, that this is how the team works now, and protects an hour a week for the champions. That one sentence outlasts three quarters of budget cuts. No sponsor, and the whole thing dies at the first bad quarter.
- What does the diagnostic actually produce?
- Six lines, plain language, no score attached: HRIS clean or scattered, process mapped or murky, three tools in play or thirteen, two curious people or none, a named sponsor or nobody, cautious culture or move fast. That's it. No slide, no maturity badge, just six honest lines you could read out loud in a stand-up in under a minute.
When reading turns into doing
The Grain Audit maps one People Ops process end to end, ranks the highest-return automations, and hands you a 90-day plan you keep whether or not we work together.
Two weeks. GBP 2,000, credited in full against a programme. Three slots a month.
Book a Grain AuditIf this resonated, there's more.
Subscribe to receive new Intelligence pieces as they're published. No noise, just the work.
By subscribing you agree to our Privacy Policy. Unsubscribe any time.
