Sample report with fictional data
AI Readiness Pulse Check
How your team sees its AI footing — and where it sees it differently.
5 participants · 1 HR, 2 IT, 1 EXECUTIVE, 1 BUSINESS
This report measures perception, not capability. Each person who took the pulse check described the organization as they see it, and the pages that follow map where those descriptions match and where they part ways. There is no score and no verdict. The pattern of agreement is the finding.
Alignment is not the ideal here; transparency and communication are. A team that agrees on everything may simply be missing information, and honest misalignment between people who see different parts of the organization can create productive tension, and sometimes a breakthrough. Differences do their damage only while they stay hidden.
So the conversations this report stages are the point, and their goal is to surface how each person reads the organization, what they want for it, and the ideas they have been carrying quietly. The goal is not to work through the differences until everyone agrees. Some gaps will close as soon as they are spoken. Others will stay open and prove useful. Both outcomes count.
One note on the charts: every dot is one person’s read. Where dots cluster, the group shares a story. Where they spread, it is telling more than one.
Start with what’s already true: AI use at Midwest Mutual is real. Adjusters are drafting correspondence with it, analysts are summarizing claims files, and at least one team has built a working experiment worth talking about. None of that waited for a strategy document, and it shows up in your results as pockets of genuine confidence.
The survey’s headline is that your five leaders describe five noticeably different organizations. Overall alignment sits at 49%, and every one of the six dimensions lands in contested territory: nobody is sure, and nobody agrees. Executives see emerging clarity; HR and operations see improvisation. IT sees usable data; the business sees locked doors.
That pattern matters more than any individual rating because it compounds. A governance question that IT thinks is settled and HR thinks is unwritten produces shadow usage. A scaling philosophy the CEO thinks is ‘prove it first’ and the claims team thinks is ‘just ship’ produces pilots that die quietly. You are not yet having one conversation about AI; you are having several, in parallel, with different vocabularies.
The good news is that this is a coordination problem, not a capability verdict. The pieces exist. What’s missing is a shared map of where you are, and the divergences in this report are the map.
The Biggest Insight
Your organization’s AI conversations are running in parallel rather than together. Where leadership sees an emerging strategy, the people closest to the work see improvisation, and that gap, not any technology deficit, is what’s holding back everything downstream.
Before any dimension detail, one fact about the five of you: the same two people sit at the edges of nearly every question. One voice consistently describes an organization finding its footing; another consistently describes one still at the starting line. The gap between those two readings, nearly two full points, is the widest thing in this report.
Anonymized: the same letter is the same person on every dimension. The gap between the outer voices is the widest thing in this report — and the first conversation.
Each dot is one person, standing where their answers put them. Position is how ready that person believes you are; how the dots group is how many stories your team is telling. Widest split first.
Every dimension has two readings: the stop people chose when asked where the organization stands, and the place their detailed answers add up to. Where the two disagree, in either direction, that gap is worth naming out loud.
The survey’s last question: “If your group could get honest about just one of these in the next 90 days, which would matter most?” The group splits 4 ways on where to start.
Built from your widest splits and your own words. Run them yourselves, or have them facilitated — either way, these are the conversations this report exists to start.
Four of you put cross-functional coordination near the starting line; one voice sees it working. That is a wide gap. Either real coordination exists that most of the room can’t see, or one person is describing something the others don’t experience.
Start here: The person who rated coordination highest: describe the coordination you see, concretely. Everyone else: say where that picture stops matching your week.
One story looks like: Everyone can name the same two coordination mechanisms and say honestly whether they touch their own work.
Two of you describe an organization with an emerging AI direction; three describe improvisation. Both camps answered the same questions about the same company.
“We have ambition but no shared language for what AI means for our people strategy.”
— one of your respondents, verbatim
Start here: Each person, one sentence: where does Midwest Mutual stand with AI today? Write the sentences side by side and mark where they diverge — that divergence is the agenda.
One story looks like: Five people give recognizably the same one-sentence answer about where the company stands.
One voice puts experimentation near the bottom while others see real motion — and the adjusters’ AI use everyone mentions is still informal. The split isn’t about whether experiments happen; it’s about whether anything comes of them.
Start here: Take the claims-file summarization already happening: what specifically stands between it and a sanctioned, supported tool? Name the decision and the decider.
One story looks like: There is a named path from pilot to production, and the current experiment is on it.
Two different companies, one org chart
Do your people know where you’re going, and does it match what’s already happening on the ground?
Your executives rate strategic clarity near 4 while HR and operations sit at 2. People know AI matters here; they disagree on whether anyone has said where it’s going. The direction may exist, but it hasn’t landed evenly.
Learning is happening, mostly off the books
Your people are already learning. Is the organization keeping up, or getting in the way?
Nobody here rates skills high, but the spread is modest; the group agrees this is unfinished. The interesting signal is in the open text: people describe colleagues teaching themselves. The organization hasn’t decided whether to fund what’s already underway.
IT sees pipes; the business sees a wall
Where is data actually blocking real use cases, and where is it a convenient excuse for inaction?
IT rates data readiness around 4. The people who would use that data rate it closer to 2. Both can be right: the infrastructure may work while access, permissions, and know-how still block real use cases.
Governance nobody can describe
Is your governance accelerating or decelerating adoption? Light-touch isn’t wrong if it’s working.
Scores scatter from 1.4 to 3.4, and the forced-choice answers suggest people genuinely don’t know what the rules are. That uncertainty pushes usage into the shadows faster than any policy could pull it back.
Six functions, six definitions of AI
Different departments mean different things by ‘AI.’ That’s a conversation to have, not just a problem to solve.
This is your lowest and most divergent dimension. Claims, IT, HR, and finance each describe a different AI conversation. Until those conversations meet, every other dimension inherits the confusion.
Pilots start; almost nothing ships
Are successful experiments spreading naturally, or dying in the pilot-to-production gap?
The group splits on whether experiments are working: IT points at pilots, operations points at the absence of anything in production. The claims-adjuster experiment everyone mentions is your test case for the pilot-to-production gap.
Each pair asks where your team thinks the organization should sit on a real tension, and whether it lives that choice today. The most divided pairs get the full picture; the rest are summarized.
When an AI pilot succeeds, which is closer to the right next step for your organization?
Scrappy-and-scaling versus prove-then-promote splits leadership from the floor. Maximum misalignment: people disagree on the right approach and on what’s actually happening today.
Scaling shows maximum misalignment: disagreement about the right approach and about current reality. With 88% of AI pilots industry-wide never reaching production, the claims-adjuster experiment is your live test of whether Midwest Mutual is the exception.
When it comes to funding AI initiatives, which is closer to the right approach for your organization?
Half the group wants strategy set at the top; half wants it to emerge from working teams. That’s a real philosophical split, not noise. Whichever side wins, nobody thinks the current approach is being executed; execution scores scatter across the full range.
The strategy pair splits the room between top-down direction and emergent, team-led discovery. Research favors the emergent camp more than most executives expect: MIT’s work puts personal-tool success rates around 40% against 5% for enterprise programs. The question isn’t which philosophy is right; it’s whether your official approach matches where your wins are actually coming from.
When it comes to building AI skills, which is closer to the right approach for your organization?
The group broadly agrees people should learn by doing, with structured support close behind. Agreement on the value, disagreement on the reality: some see learning happening, others see people left alone with a chatbot.
Everyone wants people to learn by doing, and nobody agrees whether that’s actually supported. Knowledge workers report saving 40 to 80 minutes a day with consumer AI tools; your open-text answers suggest some of that is happening here, unfunded and unnamed. Naming it may be the cheapest capability investment available to you.
Who should drive AI decisions splits the room: IT wants a center of gravity, operations wants each function running its own agenda. Both are answers to the same frustration.
Perfect-the-data versus start-messy divides the room roughly along technical lines.
Near consensus: guardrails should enable use, not gate it. This is your strongest shared value.
Nothing in these results says Midwest Mutual can’t do this; the results say you haven’t yet decided together what ‘this’ is. That’s a solvable problem, and the fact that five busy leaders answered honestly enough to expose it is the strongest signal in the data.
Your pattern is common in mid-size insurers but rarely named: grassroots adoption running ahead of shared direction, with each function narrating its own version of events. The leverage isn’t a bigger AI program; it’s getting the five people in this report seeing the same map, because every dimension you rated improves the moment the conversations merge.
If it would help to have that conversation facilitated by someone who has run it across many organizations, that’s the work Harness Intelligence does. Share this report with us and we’ll come to the discussion with your specific divergences already mapped.
Your report is shared only with Harness Intelligence. Never sold, aggregated, or shared with anyone else.