No result found
This report needs a scored result to generate.
This is a static sample and does not depend on a scored result.
Back to the course page
PRAXORALAB Illustrative example · fictional company, for format reference only
Organizational Adoption Profile Diagnostic
Portside Retail Group
Overall Adoption Score Developing
A mixed profile: four of five adoption drivers are developing well, with one named constraint, psychological safety, holding back genuine behaviour change beyond the first rollout regions.
| Driver | Score | Status | |
|---|---|---|---|
| Adaptability Mindset | 2 / 3 | Developing | |
| Psychological Safety | 1 / 3 | Blocker | |
| Empowerment Orientation | 2 / 3 | Developing | |
| Action Style | 2 / 3 | Developing | |
| Adoption Capacity | 2 / 3 | Developing |
Priority: Psychological Safety
Store staff can name at least one occasion where a colleague was corrected in front of the team for a reorder the copilot recommended and got wrong; nobody interviewed for this diagnostic could name an occasion where a manager owned an AI-driven miscall themselves, in front of the same team.
A nine-out-of-fifteen score, with four drivers clustered at "developing" and one sitting alone at "blocker," is a more useful result than a flat five-across-the-board mid-range score would have been, and it is worth being precise about why. When every driver scores the same, an organisation usually concludes it has a general maturity problem and responds with a general maturity programme, training for everyone, on everything, which is exactly the kind of well-intentioned effort that produces activity without adoption. Portside doesn't have that problem. It has one driver, Psychological Safety, sitting a full point below the other four, and that gap is doing almost all of the explanatory work for the rest of the profile.
Look at what happens once you read the other four scores next to it rather than in isolation. Adaptability Mindset scores a 2 because staff pick up the tool quickly and describe it positively, which tells you the resistance here is not a competence or attitude problem. Adoption Capacity also scores a 2, and for the same reason: Portside already runs a second AI tool successfully, so the organisational muscle to learn new systems clearly exists. Put those two together and a flat maturity score would have hidden something important, that the constraint isn't Portside's general readiness for AI, it's a specific, nameable condition in how people are treated when the tool gets something wrong.
Empowerment Orientation and Action Style, both also at 2, are where that condition becomes visible in behaviour rather than in a survey answer. A department lead who can technically override the model but leaves the reason field blank eighty percent of the time is not exercising judgement in the open, they are exercising it quietly, in a way that protects them if questioned later. That is not empowerment, whatever the survey says about it; it's a workaround, and workarounds are the behavioural signature of low psychological safety, not of low capability.
The reason I'd tell Portside's leadership to fix Psychological Safety before touching anything else is not sentiment, it's sequencing logic: every other driver on this page is already scoring in the range where the fix is refinement, not construction. Spending budget or attention on Empowerment Orientation before Psychological Safety moves is refinement work applied to a foundation that isn't set yet, and I would expect it to look like progress on a dashboard while changing very little about what actually happens on a store floor.
PRAXORALAB Naming the Resistance Pattern
Built for one live rollout brought into the room. Portside Retail Group's example rollout: an AI-assisted inventory and replenishment copilot rolled out to store operations teams across 140 UK stores. The framework distinguishes activity metrics that look healthy from adoption that has actually changed how people work, and names which pattern, if any, is present in the room's own evidence.
Diagnosis: Silent Resistance
Not Checkbox Adoption, where usage is faked to satisfy a metric while nothing about the work changes. Portside's staff use the copilot exactly as instructed, on screen, then quietly route around its output where nobody is watching, because being visibly wrong with the tool currently carries more risk than being quietly wrong without it.
Usage dashboard vs. shift-level behaviour
The rollout dashboard records 76% of the copilot's suggested reorders as "accepted" across all 140 stores, the number head office has been reporting upward as an adoption success.
Shift-level stockroom logs at nine stores sampled for this diagnostic show a different pattern: a majority of "accepted" orders were manually adjusted afterward, off-system, before being placed with the supplier.
Exit interviews
Two of six department leads who left Portside in the last quarter cited, unprompted, being afraid to get a reorder wrong in front of their team as a reason they quietly stopped relying on the copilot's suggestion before their last day.
Silence in weekly briefings
The copilot is a standing agenda item at all 140 stores' weekly operations briefings.
Store manager briefing logs show it is actually discussed, beyond a one-line status update, in only one briefing in five.
I want to be specific about why I'm calling this Silent Resistance and not Checkbox Adoption, because the two patterns look similar from a head office dashboard and require almost opposite responses. Checkbox Adoption is what you get when people go through the motions of using a tool, logging in, clicking accept, because usage is measured and they know it, while genuinely changing nothing about how the work gets done; the tell is usually a suspiciously uniform, evenly-distributed usage number across a whole organisation. Silent Resistance is what you get when people quietly route around a tool specifically because they've learned that being wrong with it is more exposed than being wrong the old way; the tell is a gap between what the dashboard reports and what the shift logs show, which is precisely what the first piece of evidence on this page contains.
Seventy-six percent "accepted" against a majority of those same orders being manually adjusted off-system afterward is not a rounding error or a data-quality issue, it's a working population telling you, through their actions rather than a survey answer, that accepting the suggestion on-screen is safer than contesting it out loud, and adjusting it quietly afterward is safer than being seen to distrust the tool. That is a textbook Silent Resistance signature, not a Checkbox Adoption one; nobody at Portside is faking usage to look good, they're using the tool exactly as instructed and then unwinding the result where nobody's watching.
The exit-interview evidence is the piece I'd push hardest on in a real engagement, because two departing leads volunteering the same reason, unprompted, in exit interviews that were not designed to surface it, is a stronger signal than almost anything a satisfaction survey could produce. People rarely volunteer their real reason for disengaging with a new system while they still work somewhere; they volunteer it on the way out, when there's no longer a cost to being honest. That two independent people gave the same answer, in the same quarter, is not a coincidence worth dismissing.
The one-in-five briefing statistic closes the loop: the tool is formally on every agenda and substantively discussed at barely a fifth of them. That is not a communication failure, a better agenda template will not fix it. It is what Silent Resistance looks like at the level of a meeting: technically present, actually avoided, because raising a real disagreement about the tool in front of the room carries more risk than staying quiet does.
Portside Retail Group, its rollout and every figure quoted above are invented for this sample only, to show the shape of the output, not a real client's actual data. In the session, this diagnosis is named from the room's own evidence, not assigned from a template.
PRAXORALAB Prioritised Roadmap
The scorecard and named diagnosis above, brought together into one prioritised plan: which driver to fix first, which to deliberately leave alone, and what it means for the rollout Portside Retail Group runs next.
- Psychological Safety, the diagnostic's weakest driver at 1 out of 3, because it is the root cause behind the Empowerment Orientation and Action Style gaps above it, not a parallel, independent problem.
- Retire the override-reason field's optional status; make a one-line reason mandatory, and visibly used for coaching, never for blame, starting with the next region's launch.
- Put one instance of a manager publicly owning their own AI-related miscall on the agenda of the next regional briefing, before asking staff to do the same.
- Adoption Capacity. Store operations already run a second AI-assisted tool successfully; another skills-building programme here would spend budget solving a problem Portside doesn't have.
- The copilot's underlying reorder-suggestion accuracy. The evidence points to a trust and safety gap in how suggestions are acted on, not to the model's own recommendations being wrong.
- Portside's next planned AI tool, a merchandising-planning assistant for regional buyers, should not launch region-by-region on the same cadence as this one did; psychological safety needs to be visibly established in at least one region first.
- The override-reason field's redesign, once fixed here, should ship as a template for every future tool's own override or exception path, not be rebuilt from scratch next time.
The instruction to leave Adoption Capacity alone is the part of this page I most want Portside's leadership to actually follow, because the instinct in most organisations facing a 9-out-of-15 score is to spread the fix evenly across every driver that isn't a 3, on the theory that more attention everywhere can't hurt. It can. Every hour a regional operations director spends running an Adoption Capacity workshop for a store team that has already proven, with a second live AI tool, that it can absorb new systems, is an hour not spent on the one driver actually holding this rollout back. Fixing what isn't broken isn't neutral, it's a real cost, paid in the attention and credibility leadership has left to spend on the driver that is broken.
The sequencing logic behind Psychological Safety going first is the same one I use with every client organisation I diagnose against this framework, not something specific to retail: a driver that scores low because of a structural or behavioural condition, not a skills gap, has to move before the drivers built on top of it can move honestly. Empowerment Orientation's blank override-reason fields will not fix themselves by making the field mandatory alone, a policy change without a safety change just produces reasons written to sound safe rather than reasons written to be true. That's why the mandatory-reason-field fix on this page is listed inside the Psychological Safety block, not as its own separate line item; it only works once leads believe an honest reason won't be used against them.
The recommendation for the next rollout, staggering region launches so Psychological Safety is visibly established somewhere before the merchandising-planning tool goes wide, is the highest-leverage line on this entire page, and it's the one I'd expect to get cut first under a deadline. It is much easier to defend "we're rolling out on schedule" to a steering committee than "we're deliberately slowing down region two until region one shows the pattern has actually changed." I would rather Portside defend the slower schedule now than diagnose the same Silent Resistance pattern again, eighteen months from now, on a second tool, having already had the evidence to prevent it.
PRAXORALAB A Concrete First Move
The one action the diagnostic on page one points to first, specific enough to run in the next two weeks, not a general principle about building trust.
Once every two weeks, each store's department lead runs a 15-minute review of that store's copilot overrides from the shift logs, out loud, with the team present. The rule that makes it work: the lead states their own most recent override first, including one that turned out to be wrong, before asking anyone else to explain theirs. The review's only output is a shared note of what the copilot missed, not a name attached to a mistake.
This is deliberately not a training module or a policy memo. Psychological Safety scored a 1 because staff have seen correction happen in public and ownership stay private; the fastest way to reverse that pattern is a manager repeatedly modelling the opposite, in the same room, on a fixed cadence, until it stops being notable.
The reason I specified who runs this move, the department lead, not head office, and not HR, is the single most important design decision on this page, more important than the fifteen-minute duration or the two-week cadence. Psychological Safety is not damaged by an absence of policy at Portside; there's no shortage of memos telling staff the copilot is meant to be a helpful assistant, not a scorecard. It's damaged by a lived pattern of who gets corrected in public and who doesn't, and lived patterns are only reversed by lived counter-examples, delivered by the same people who created the pattern in the first place. A head-office-run session, however well designed, would be received as one more thing being done to store teams rather than something their own manager is doing differently.
The instruction that the lead states their own most recent override first, including one that was wrong, is doing more work than it looks like on the page. It converts the session from a review of the team's mistakes, which is what "override review" sounds like on paper, into a demonstration that admitting an AI-related miscall out loud doesn't cost the person admitting it anything. That demonstration has to come before anyone asks a department associate to do the same thing, not alongside it, because asking first and modelling second is functionally indistinguishable, from the team's point of view, from every other well-intentioned initiative that quietly punished the first person honest enough to test it.
I'd flag one risk in a real engagement that this page doesn't have room to fully draw out: a blameless review only stays blameless if it survives the first genuinely costly mistake that surfaces in it, a wrong reorder that visibly affects revenue or a customer, not a minor one. The test of whether Portside has actually built psychological safety, rather than just started a well-run meeting, is what the department lead does in the room the first time an override review surfaces something that stings. If the response in that moment is still curiosity about what the copilot missed rather than a search for who's at fault, the pattern has genuinely shifted. If it isn't, the fifteen-minute ritual will have taught the room the opposite lesson faster than any single incident would have on its own.
Every score, quote and figure on these four pages is invented for Portside Retail Group, a fictional company, so the format of what a participant leaves with can be judged before enquiring. It is not a real client's deliverable, and no organisation named Portside Retail Group is a Praxora Lab client. The session itself scores your own rollout's own data, in the room, on the day.
3.5 hours, one session, facilitated by Dr. Xenia Wade.