- Apr 27
JERK Report #11: The Debate Is Over
98% of firms are testing AI. Only 5% report bottom-line gains. Here’s what the winners do
You bought the tools. That was the easy part.
The IMF added AI adoption to its macroeconomic dashboard this week.
That means the debate is over.
Not the debate about whether AI is useful. The debate about whether you have time to figure it out.
You don't.
The question now is how to navigate it before the gap between experimentation and real performance gets wider.
Position — where things stand right now
Three numbers, as of Q1 2026. Hold them together.
98% of firms are testing AI. Only 5% report meaningful bottom-line gains.
97% of surveyed knowledge workers are using AI poorly or not at all.
74% of AI’s economic value is being captured by just 20% of organizations.
Adoption is nearly universal.
Transformation is rare.
Velocity — which direction is it moving?
The bottleneck is not the technology.
BCG is direct: success is 70% people and ways of working, 30% technology.
Chip and Dan Heath wrote about why this is hard in Switch. Change fails not because people do not understand the case for it. It fails because they do not feel it.
You can make the rational argument all day. Without the team feeling it, nothing moves.
If you have been giving permission for AI use and wondering why nothing is compounding, you are not behind. You just have not been asked the right question yet.
Acceleration — What’s being measured wrong
This is where most organizations lose the signal.
Employees who complete AI training still score an average of 40 out of 100 on proficiency. Still experimenter territory.
Most organizations measure success by login rates and seat counts.
These are trailing indicators. They tell you what already happened.
The leading indicators look different. Not accounts activated, but AI workflows actually deployed, built into recurring work. Not prompts sent, but problems solved repeatedly, week after week.
And the single metric that predicts almost everything else:
Employees whose managers actively expect AI use are 2.6x more proficient than those whose managers do not.
Not manager permission.
Manager expectation.
This is not really a measurement of AI fluency. It is a measurement of whether the team has been told what is expected of them.
Who is winning — and what they do differently
The winning organizations are not further along the same path. They are on a different path entirely.
Grant Thornton’s 2026 survey of 950 business leaders found that organizations with fully integrated AI report revenue growth at 58%, while those still piloting report 15%.
The causality runs both ways. Companies already growing have more capacity to integrate. But the pattern is consistent. The piloting organizations are not behind because they started later. They are behind because they are doing more — more pilots, more use cases, more functions touched — without the discipline to know what is working.
The organizations pulling ahead are scaling fewer initiatives, not more. With better measurement. Clearer exit criteria. Depth over breadth.
But focused application is more than the number of use cases. It is also the type.
PwC found the wider pattern. The top 20% are not using AI only to reduce costs. They are using it to pursue growth: new revenue, new business models, new markets.
They point AI at what matters.
The companies winning are not smarter.
They are more deliberate.
Jerk — The signal before the signal
Two things shifted this quarter that matter more than the usual hype cycle.
First, AI adoption is now being tracked more explicitly by major economic institutions, which means the measurement stack is finally catching up to reality.
That is not a velocity change. AI adoption has been rising for three years. That is a change in how seriously the infrastructure around it is taking it.
The second jerk signal is closer to home.
The 2.6x gap between expectation-led teams and permission-only teams is the current acceleration reading. The jerk is whether that gap bends over the next four quarters.
It will either close, as expectation becomes the norm and the multiplier shrinks. Or it will widen, as the teams already pulling ahead compound their advantage.
The data does not yet tell us which.
My hunch is that it widens.
The teams already running expectation-led practice are building organizational muscle that compounds weekly. The teams stuck at permission are still negotiating whether AI use is appropriate. Those two trajectories do not converge on their own. Something has to interrupt them.
If the gap widens, the cost of being a permission-only manager goes up. If it narrows, the window for easy advantage closes.
Either way, the read is the same. The acceleration of AI capability is not the variable that matters most right now. The acceleration of who is expected to use it is.
That is what the institutions are starting to measure.
That is what your team is already living.
Your five-minute practice this week- Find your pivotal use case
Open your LLM of choice. Paste in this entire Jerk Report and then paste this prompt below:
Using the framework above, help me find my highest-leverage AI use case. I'll list five recurring tasks I do most weeks. Apply the two filters with me: which would change the shape of my work, and which touches growth. Then help me write two sentences I can say to my team to make this expectation visible.
The 2.6x multiplier starts with one person making expectation visible. That person can be you.
The JERK Report is a weekly signal read for small business owners. One signal. Four layers. A five-minute practice. Every Monday.
From Rose Thun at Design Rosetta.
Sources
The State of AI: Global Survey 2025
https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-aiBenchmarking Section’s AI Proficiency
https://www.sectionai.com/ai/the-ai-proficiency-reportPwC’s AI performance study
https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-ai-performance-study.htmlAI Transformation Is a Workforce Transformation
https://www.bcg.com/publications/2026/ai-transformation-is-a-workforce-transformationAI Adoption Metrics
https://zapier.com/blog/ai-adoption-metrics/Grant Thornton 2026 AI Impact Survey
https://www.grantthornton.com/services/advisory-services/artificial-intelligence/2026-ai-impact-surveyMonitoring AI Adoption in the US Economy
https://www.federalreserve.gov/econres/notes/feds-notes/monitoring-ai-adoption-in-the-u-s-economy-20260403.html
Check out the Jerk Report,
The JERK Report is a weekly signal read for small business owners. One signal. Four layers. A five-minute practice. Every Monday. From Rose Thun at Design Rosetta