AI-native roadmap
Where Does Your Team Actually Stand on the AI-Native Path?
Planning season already asks "where do we stand." Here's the version of that question worth asking about AI-native maturity, honestly.

Planning season already forces a version of this question. Budget conversations ask where the team stands against its goals. Roadmap reviews ask what’s actually landed versus what’s still aspirational. The same honest placement is worth doing for AI-native maturity specifically, and planning season is exactly the moment most leaders are already in the right headspace to do it.
A planning-season health check means honestly placing your team on the AI-native path, not by counting which AI tools you’ve bought, but by checking whether agents and people work from the same current source, and whether that holds at your team’s actual size. That’s a narrower, more useful question than “are we AI-native yet,” and it’s worth answering deliberately instead of leaving it to a vague sense of how things are going.
Where does your team actually stand on the AI-native path, right now?
Honestly, and specifically, not by tallying tool subscriptions or comparing yourself to whatever the loudest company in the space claims to be doing. The real placement comes from checking a handful of concrete things: whether agents work from a shared, checked source instead of each pulling its own picture; whether there’s a genuine approval gate rather than blanket auto-approval or blanket manual review; and whether status and specs stay current without someone manually re-syncing them by hand. Answering those honestly usually produces a mixed picture, further along in some places than a team would guess, behind in others, which is a much more useful result than a single yes-or-no verdict about whether the team has “arrived.”
Why is planning season specifically the right moment for this?
Because the budget and roadmap conversations already happening ask a version of the exact same question this health check answers, just about revenue or delivery instead of AI-native maturity. Planning season is when a team is already assembling an honest account of where it stands against its goals for the next cycle, comparing what was planned to what actually happened, and deciding what to prioritize next. Folding an AI-native placement into that same moment costs little extra, because the muscle of honest assessment is already engaged. Doing it at a random point mid-quarter means starting that reflective work from scratch, competing with whatever else is actually on fire that week.
What does “good” actually look like this year, concretely?
Not “adopt more AI tools.” A shared, checked source that both people and agents work from, a real tiered approval gate instead of an all-or-nothing rule, and status and specs that stay current without someone manually re-syncing them, the same three markers that separate genuine AI-native maturity from AI adoption dressed up as the same thing. Those three, plus how fast a dropped handoff gets caught and how quickly new context spreads to a new team member or a new agent, are the concrete signs worth checking a team against, not a vague sense of how many AI vendors are on the invoice list this quarter.
That specificity matters because “good” keeps shifting as a vague target. Last year’s version of AI-native adoption looked like having any coding assistant at all. This year’s version assumes that baseline and asks a harder question: does the team have a real structure underneath the tools, or did the tools get added on top of an unchanged process that would keep running the same way if they disappeared tomorrow. Planning season is the moment to check which of those two describes your team honestly, not the moment to restate last year’s goal as if it still counts as ambitious.
What should the assessment itself actually produce?
A clear placement, specific about where the team is further along than assumed and where it’s genuinely behind, not a single pass-or-fail score that flattens a mixed picture into one discouraging number. A team that’s strong on governance and weak on living documentation needs a different next step than a team that’s the reverse, and a single overall percentage tells neither team which gap to close first. The value of an honest assessment is in the shape of the result, not a grade.
The same team, one planning season later
Take the fifteen-person team from an earlier honest scorecard: one clear yes on approval gates, two partials, two clear nos, including a living-documentation gap nobody had closed. Planning season is where that placement gets revisited, not just filed away. They come into the session expecting to still be behind across the board, since that’s the vague feeling most leaders carry in. The actual re-check comes back more specific than that. Approval gates are still a clear yes, the same real tiered rule, requiring a person on ownership changes and priority flips while letting small ticket updates clear automatically. Dropped-handoff catching has moved from partial to more consistent: the senior engineer’s habit is still doing real work, but it’s no longer the only thing catching a stall. Living documentation, the one clear no from before, hasn’t moved: specs still get written once at kickoff and rarely revisited, and engineers are still keeping private notes on what actually changed since.
That’s a placement, not a score, and it’s also a trend, not just a snapshot: two signs holding steady or improving, one gap that planning season now shows has gone unaddressed for a full cycle. The team walks away with a sharper next step than before, closing the living-documentation gap before the next planning season rolls around, rather than a vague resolution to “be more AI-native this year,” which isn’t a plan anyone could actually execute against.
Why does tying this to concrete signs matter more than a general impression?
Because a general impression of “we’re doing okay” or “we’re behind” doesn’t tell anyone what to actually do differently next quarter, while a placement against specific signs does. The same scaling reality that applies as a team grows from ten engineers to a hundred applies here too: an informal sense of how things are going works fine at small scale, when a leader can see most of what’s happening firsthand, and it stops being reliable the moment the team, the agent count, or the surface area of work grows past what one person can informally track. Planning season is exactly the point in the year when that informal sense gets tested against reality anyway, through budget numbers and delivery metrics. Testing the AI-native placement the same way, against specific signs rather than a gut feeling, is the same discipline applied to a newer question.
Google’s own internal guidance on running structured goal-setting cycles makes a related point directly: periodic, honest calibration against specific criteria is what makes an annual or quarterly planning rhythm actually useful, rather than a ritual that restates the same vague intentions every cycle without checking them against anything concrete. An AI-native health check folded into planning season is that same calibration, aimed at a newer, faster-moving kind of goal.
What’s the natural next step once a team has this placement?
A proper AI-native roadmap assessment, which goes deeper than a planning-season gut check into exactly where a team stands and what the specific next move should be. The health check described here is meant to be fast: something a leader can walk through honestly in the same session where they’re already reviewing the quarter’s numbers. The fuller assessment is the place to take a placement that’s flagged a real gap, like the living-documentation example above, and turn it into an actual plan, not just an accurate diagnosis. Getting a clear sense of where a team stands on the broader AI-native path is the starting point either way; planning season is simply the moment most leaders are already positioned to do it honestly.
FAQ
Does this replace the regular planning-season review, or run alongside it?
Alongside it. The AI-native placement is a specific addition to the same reflective moment, not a separate process competing for a different meeting.
What if the honest placement comes back worse than expected in every area?
That’s still a more useful result than a vague sense of falling behind, because it names specific gaps a team can actually prioritize, rather than leaving “AI-native” as an undifferentiated goal for next year too.
How is this different from just counting which AI tools the team has adopted?
Tool count says nothing about whether the workflow actually changed around those tools. The signs this health check checks, a shared source, a real approval gate, current specs, measure structure, not subscription count.
Does a small team need to worry about this as much as a larger one?
The signs apply at any size, but the cost of an honest gut check is lower for a smaller team precisely because the gaps are still cheap to close. Waiting until growth makes an informal approach unreliable is the more expensive path.
Is there a pass/fail threshold this health check is building toward?
No. The goal is an accurate placement, not a score. A team further along on some signs and behind on others is a normal, useful result, not a failing one.
Curious where these signs would actually place your own team this planning season? See what an AI-native roadmap assessment would surface.
Or skip the self-assessment: try Thor free and see what it picks up from your own Slack and tracker.