Research
Somebody has to keep score. Nobody was before.
Roughly nine in ten companies use AI. About three in ten can show a return. Every number in between comes from a vendor with something to sell. AiAi measures who is getting a return from AI, how, and what separates them from everyone else, and we publish what we find whether or not anyone likes it.
~90%
of companies use AI
~29%
see a significant return
Industry survey figures, 2025–26. Our own benchmark replaces these with primary data in 2027. [citations: Becky]
What we measure
Two instruments. One question: is it paying off?
The metric
ROAI, Return on AI
A documented business result attributable to AI, set against what was spent to get it. Hours reclaimed, cycle time cut, errors removed, revenue moved. Adoption is proven by the result, never by what was deployed or what the invoice said.
Every member company establishes a baseline, works problems, and reports movement. The aggregate, by function, company size and industry, is the benchmark everyone else gets measured against.
The scale
The Operational Maturity Scale
Five levels: Awareness, Experimentation, Integration, Strategic Scaling, AI-Native. It places a function, or a whole company, and tells you whether it's moving. Most companies sit on level two and have for a while.
1
Awareness
2
Experiment
3
Integration
4
Scaling
5
AI-Native
The research program
What we're publishing, and when.
We'd rather tell you what's coming than pretend it's already here. First editions land in 2027 as the first circles produce data. Sign up below and you'll get each one the day it's out.
Flagship · annual
The State of AI Adoption
The annual report on who is getting a return from AI in mid-market companies (50–5,000 employees) and what they do differently. Where each function sits on the Operational Maturity Scale, what moved in a year, where the money went and where it came back. Unveiled at the annual summit. Headline findings are public; the full report goes to members, their CEOs and research partners.
Quarterly
The ROAI Benchmark
Where return is showing up this quarter, by function and company size, drawn from member assessments and quarterly check-ins. The public cut is a two-page brief. Members get the full tables and their own position in them.
Quarterly
The CEO's AI Confidence Pulse
A short survey of mid-market CEOs: how confident are you that your AI spend is paying off, and what would change your mind? Built on the two thousand CEO conversations that started AiAi, now put on a repeatable footing.
Ongoing · by function
Function Briefs
What's actually working in HR, Finance, Operations, Marketing and Sales right now, with the problem, the approach and the measured result. Drawn from what circles work on each month, anonymized. The living version is the Benchmark Vault, for members.
Ongoing · with partners
Special Studies
Deeper questions worked with academic partners: why pilots stall between level two and three, what IT ownership does to adoption, how company size changes the playbook. Developed with university research partners, beginning in Denver.
Who gets what
The headline is free. The depth is for the people in the data.
Members contribute the data, so members and their CEOs get the deeper cuts. Everyone else gets the findings that matter most, free, because the point of keeping score is that people see it.
How the data is gathered
Primary data from companies doing the work.
Every member completes the ROAI Assessment on joining and a check-in each quarter. Circles log what they work on and what happened. CEO surveys run alongside. No scraped vendor case studies, no self-selected webinar polls.
Independence
Sponsors fund how far it travels. Never what it says.
We report what the data shows, including when it's unflattering to a member, a sponsor or a popular tool. Sponsors never see member data and never review findings before publication. If that's a problem, we're the wrong research partner.
Confidentiality
Your number is yours. Only the aggregate is published.
Company results are confidential to the company, its sponsor and its Catalyst. Published work uses aggregates with minimum cell sizes so no company can be picked out. Methodology is led by Dr. Rebecca Nusbaum, Head of Research & Education, and published with each report.
The ROAI Assessment
Find out where your function sits on the scale. Ten minutes.
The same instrument members use for their baseline, self-scored. You get your level, what the next level looks like, and how companies your size compare. Opening early 2027; sign up below and you'll be first.
Get the research
Every report, the day it's out. Free.
Pick what you want. We'll send the public cut of each one, plus the newsletter if you ask for it. Nothing else, and we'll reply personally if you write back.
Want to be in the research, not just read it?
Tick the research panel box and we'll invite you to the CEO's AI Confidence Pulse and special studies. Panel members see results before the public does.
Universities and sponsors
We partner with university researchers in every city we enter, and we accept sponsorship for distribution under the terms above. Write to research@theaiai.net.
Want your company in the data, with a room to act on it?
Members get the full research, their own quarterly ROAI report, and a room of peers to work the problems with.