--- title: "AI adoption interview guide for employees | Spradley" description: "A ready-to-use 25-minute AI adoption interview guide: who to interview, questions and probes built around real tasks, and how to analyse the answers." url: https://spradley.io/articles/ai-adoption-interview-guide --- # AI adoption interview guide for employees A ready-to-use 25-minute AI adoption interview guide: who to interview, questions and probes built around real tasks, and how to analyse the answers. Seb Dackus Co-founder & CPO Jun 18, 2026 ![A younger colleague crouches beside an older woman at her desk and points towards a computer](https://spradley.io/media/blog/ai-adoption-interview-guide/hero.webp) This AI adoption interview guide finds out why people use, avoid or quietly work around the AI tools you rolled out, by walking through the last real task they used them for. It takes about 25 minutes and has seven sections with open questions and follow-up probes. Use it four to eight weeks after launch, when people have tried the tools on real work, and again a quarter later. Section Time What you learn 1\. Opening 2 min Trust, ground rules 2\. Their work 3 min Where AI could fit 3\. The last time 6 min Actual use, step by step 4\. When it didn’t work 4 min Why people stop 5\. What’s allowed 3 min Policy clarity, quiet use 6\. Manager and team 5 min Support, norms, worries 7\. Close 2 min What would help ## Who should you interview about AI adoption, and when? Interview three groups: heavy users, people who tried the tools and stopped, and people who never started. Licence data tells you which is which. The middle group usually teaches you most, because they can tell you the exact moment it stopped being worth it. Cover the frontline as well as the office. BCG’s [AI at Work 2025](https://www.bcg.com/publications/2025/ai-at-work-momentum-builds-but-gaps-remain) research found regular use among frontline employees had stalled at 51%, that regular use was sharply higher among people who had at least five hours of training, and that only about a quarter of frontline employees said they got strong leadership support. A guide that only reaches head office will miss that gap. Interview managers too, with the same questions plus section 6 from their side. [Gallup’s AI indicator](https://www.gallup.com/699797/indicator-artificial-intelligence.aspx) (data as of May 2026) calls manager support the strongest driver of widespread adoption: employees whose managers actively support AI use were 1.7 times as likely to use it a few times a week or more. ## The AI adoption interview guide If you already run a survey, our list of [AI adoption survey questions](https://spradley.io/articles/ai-adoption-survey-questions) covers breadth. This guide goes deep on fewer topics. The core of it is section 3: one real task, walked through step by step. People are poor at summarising their own habits and good at describing what they did on Tuesday. ### 1\. Opening (2 minutes) 1. What is your role, and what does a typical day involve? Before this, say plainly that the interview is about the tools and the rollout, not their performance, and that using AI or not using it are both fine answers. Without that, people who use unapproved tools will say nothing. ### 2\. Their work (3 minutes) 1. Which tasks take up most of your week? 2. Which of those would you happily hand to someone else? Follow up with: - What makes that task tedious or slow? ### 3\. The last time (6 minutes) 1. Think of the last time you used an AI tool for work. What was the task? 2. Walk me through it. What did you type or ask, and what came back? 3. What did you do with the result? Follow up with: - How much did you change before using it? - Compared with doing it the old way, was that faster, slower or about the same? - Which tool was it, and is it one the company provides? If they have never used one, ask about the last time they thought about it and decided not to. ### 4\. When it didn’t work (4 minutes) 1. Tell me about a time it gave you something you couldn’t use. What happened? 2. Is there a task you tried once and then went back to doing by hand? Follow up with: - What would it have needed to get right for you to keep going? - Did you tell anyone about it? ### 5\. What’s allowed (3 minutes) 1. As you understand it, what are you allowed to use AI for here? 2. Has there been a moment where you weren’t sure whether something was allowed? What did you do? Follow up with: - Where did you learn the rules: training, your manager, colleagues? ### 6\. Manager and team (5 minutes) 1. How does your manager talk about AI tools? 2. Who in your team is good with them, and what do they do differently? 3. Would you be comfortable saying you used AI on an important piece of work? Follow up with: - Can you give me an example of something your manager said or did? - What would make you more or less comfortable saying so? ### 7\. Close (2 minutes) 1. What would make you use it for one more task next month? 2. Is there anything we haven’t asked that we should know? ## How do you run it: a person or an AI interviewer? A person is a good choice for ten or twenty interviews with a pilot group. A researcher can ask to see the actual prompt and spot the moment someone hesitates. It is slow and expensive to repeat across a company of several thousand, and people may play down unapproved tool use in front of someone from IT. An AI interviewer runs the same guide with everyone. Spradley does this as one-on-one interviews by text or voice. Invitations go by email or Slack, people answer on their own phone or computer in their own time, and the interviewer asks a follow-up on each answer, so the walk-through in section 3 gets the detail it needs. There is a certain irony in using AI to ask about AI, and some people will mention it. Treat that as data too. Our [AI adoption page](https://spradley.io/solutions/ai-adoption) shows how teams set it up. ## How do you analyse AI adoption interviews into themes? Start with section 3, because it is the most concrete. For each interview, note the task, the tool, whether it saved time and what happened to the output. Then group the reasons people stop or never start. The same handful tends to recur: the tool doesn’t fit the task, a bad first result, unclear rules, no time to learn, or a manager who is lukewarm. Our post on [why employees don’t use AI tools](https://spradley.io/articles/why-employees-dont-use-ai-tools) describes each. 1. Code each interview by user type: regular, lapsed, never started. 2. List the tasks people use AI for, by team. This becomes your use-case library. 3. Group the stopping reasons into themes, with two or three quotes each. 4. Compare teams. A contact centre and a legal team will have different blockers. 5. Check what managers said in section 6 against what their teams said. Report only for groups of five or more, and never pass on who admitted to unapproved tools. Spradley never shows numbers or ratings for fewer than five people. ## Adapt it - For a single tool rollout, such as Copilot, name it in sections 3 to 5 and add licence data to section 2. - For frontline staff, cut section 2 and keep the interview to 10 to 12 minutes. - For managers, add: “What do you expect from your team on AI, and have you said so?” - For the second round, repeat section 3 exactly so you can see whether the tasks have changed. ## Frequently asked questions ### How long should an AI adoption interview be? About 25 minutes for office staff. A 10 to 12 minute version that keeps the walk-through of the last task works for frontline staff. ### Should we interview people who don't use the tools? Yes. People who tried and stopped, and people who never started, explain most of the gap in adoption. ### How is this different from an AI adoption survey? A survey covers many topics briefly. This guide goes deep on one real task per person, with follow-up questions, so you learn why use stalls. ### Will people admit to using unapproved AI tools? More often when you say up front that both using and not using AI are fine, and when results are only shown for groups large enough that no one can be identified. ## Key terms Task walk-through An interview technique where the person describes one recent real task step by step, instead of summarising their habits in general. Lapsed user Someone who tried a tool on real work and then stopped using it. Shadow AI Use of AI tools at work that the organisation has not approved or does not know about. Use-case library A list of real tasks people in each team use AI for, collected from interviews and shared so others can try them. AI-moderated interview A one-on-one interview in which software asks the questions and follow-ups, by text or voice, and the answers are analysed afterwards. ## More articles [Explore all resources](https://spradley.io/articles) [ ![A man carrying a box of desk belongings stops by the lift to talk with a colleague](https://spradley.io/media/blog/exit-interview-questions/hero.webp) Interview guide ### Exit interview questions and a guide to run them Exit interview questions in a ready-to-run 25-minute guide: probes, who should ask, when to hold it, and how to sort answers into preventable turnover. ](https://spradley.io/articles/exit-interview-questions)[ ![A tired nurse resting on a staff-room sofa while a colleague leans in to talk](https://spradley.io/media/blog/burnout-check-in-questions/hero.webp) Interview guide ### Burnout check-in questions for managers Burnout check-in questions for managers, built on WHO’s definition and HSE’s six stress factors, plus early team signs to watch and where to refer. ](https://spradley.io/articles/burnout-check-in-questions)[ ![An older worker showing a new starter in stiff overalls to his locker](https://spradley.io/media/blog/onboarding-check-in-questions-30-60-90-days/hero.webp) Interview guide ### 30, 60, 90 day check-in questions for new hires Day 30, 60 and 90 check-in questions for new hires, with probes, who should ask each one, and how to read the answers across a whole hiring cohort. ](https://spradley.io/articles/onboarding-check-in-questions-30-60-90-days)[ ![A manager in a suit listening to two warehouse staff across a canteen table](https://spradley.io/media/blog/skip-level-meeting-questions/hero.webp) Interview guide ### Skip-level meeting questions and how to run them Skip-level meeting questions in a 30-minute guide with probes, how to avoid undermining the manager, group or one-to-one, and what to do next. ](https://spradley.io/articles/skip-level-meeting-questions)[ ![Office workers carrying chairs and boxes past a pinboard where blank cards are being rearranged](https://spradley.io/media/blog/post-reorganisation-interview-guide/hero.webp) Interview guide ### Questions to ask employees after a reorganisation A 25-minute interview guide for 4 to 12 weeks after a restructure: role clarity, reporting lines, decision rights, workload, lost knowledge and trust. ](https://spradley.io/articles/post-reorganisation-interview-guide)[ ![A woman reaching for her coat at the end of the day turns back to talk to her manager](https://spradley.io/media/blog/stay-interview-questions/hero.webp) Interview guide ### Stay interview questions and a guide to run them Stay interview questions in a ready-to-use 20-minute guide: who to interview and when, questions and probes, who should ask, and how to find themes. ](https://spradley.io/articles/stay-interview-questions)