A short, plain-language survey to find out where your team really is with AI — and what to do next. Two things make it work: open with how people feel, not what they know, and decide what you'll do with every answer before you send it. Here's how we run one, with two versions to start from.
There are two versions of the survey — same idea, tuned to your stage. Before anything else, pick the one that sounds like your organization right now.
For organizations that haven't formally adopted AI tools yet. Lighter on jargon, focused on awareness, appetite, and any quiet experimentation already happening under the radar.
For organizations already using AI chat tools. Goes deeper on how people use AI, what they've built, and where it breaks down.
The point isn't a score — it's knowing who needs what kind of support, and being able to re-measure later.
Choose the version above that matches where you are today — based on whether AI tools are formally rolled out yet. Send only one; mixing them muddies your baseline.
Ask all staff to respond, not just the eager ones. Partial responses give you a partial — and misleading — picture of where the team really stands.
Use the answers to sort people into the right sessions, shape demos around real tasks, and set a baseline you can re-run in a few months.
Anonymous or named? To track how individuals grow over time you'll need names attached — but say so clearly, because it changes how honestly people answer.
Know what each answer triggers. Decide what you'll do with every response before you send it. A survey that doesn't change what happens next is just a poll.
Drawn from what we've seen run well in workshops, plus what the research backs up. Each one is built to get an honest picture instead of a flattering one.
People are bad at rating themselves — so don't make them guess. The biggest risk in any maturity survey is self-rating: someone who only summarizes emails may call themselves "intermediate" because they don't know what good looks like. Every principle below is built to get past that and surface what people actually do.
Open with a quick emotional check-in — excited, curious, skeptical, nervous. People can't engage with AI until they name where they stand.
Don't say "rate yourself." Ask people to describe something they actually built with AI — a real example beats a self-chosen level.
"I use it daily" means little if it's one task. Ask what people use AI for, and in how many ways — breadth is the real signal.
Ask if people can tell a prompt from an assistant, an agent, and an automation. A gentle, revealing check on real fluency.
"Where has AI not worked for you?" shows who's really pushing the tools — and exactly where people need better support.
Open-ended "what would you automate?" builds a backlog of real ideas and makes staff feel heard going into training.
Short questions, everyday words, no acronyms in the starter version. The easier to answer, the more honest the responses.
Plan to re-run the same survey in a few months. Comparing results proves training worked — and shows where to focus next.
We'll share both versions, help you pick the right one, and turn the responses into a training plan — who needs what, and where the quick wins are.