You want a quick read on what your team, your customers, or your network actually think about something. Not a 1-to-5 score. Not a click on a poll option. An actual answer, in their own words. There has never been a simple tool for that, the way Doodle is the simple tool for "what time works for everyone" and Google Forms is the simple tool for "how many people prefer option A." Skimle Ask is built to be that tool for opinions: a chat-based AI interviewer that anyone can set up in minutes, share as a link, and get real answers back from, already organised into themes.
Why this suddenly has a name: "qual at scale"
In December 2025, Anthropic, the company behind Claude, did something that put this category on a lot of people's radar at once. They built a tool called Anthropic Interviewer, powered by Claude, and used it to interview over 80,000 Claude users, in 159 countries and 70 languages, about how they actually use AI, what they hope it can do, and what worries them. Anthropic called it the largest and most multilingual qualitative study ever run. A separate pilot of the same tool, with 1,250 professionals, found that 97.6% of participants rated the experience a 5 or higher out of 7, and 99.12% said they would recommend the format.
That story is why phrases like "qual at scale" and "AI interviewer" started showing up in conversations they had not been part of before. The pitch is straightforward: a structured AI conversation can collect real, open-ended answers from thousands of people in the time it would take a human researcher to schedule a few dozen interviews.
Some qualitative researchers were quick to point out at using AI handle to the interviewing and doing it via a short text-based discussion is not the same as having a skilled human researcher spend real time with the participants to ask thoughtful questions... and we agree. At Skimle we do not consider AI-assisted interviewing, or "qual at scale" to be a replacement for classic qualitative researcher, rather we see it as a new tool opening new possibilities.
You do not need Anthropic's research team, or a model lab's infrastructure, to use this new tool. That is exactly what Skimle Ask is for: the same basic concept, sized for a team, a customer base, or a community, not a global Claude user base.
What Skimle Ask is
Skimle Ask is a short AI-conducted conversation that anyone can launch. You write down what you want to find out, Skimle Ask turns it into a small set of initial questions, and you get a link. Anyone you send the link to opens it, chats with the AI for a few minutes, and that is the whole experience for them, no account, no app, no scheduling.
You can test Skimle Ask by answering this hypothetical example employee experience Ask.
Behind the scenes, Skimle Ask uses an AI system to synthetise user answers, ask smart follow-up questions and manage time so that you get real insights instead of just open text field type responses. Also, every response feeds straight into Skimle's analysis engine, the same one used for full qualitative research projects. So instead of getting back a spreadsheet of free-text answers nobody has time to read, you get a structured set of themes, each one linked back to exactly who said what.
It sits in a different spot to the tools you already reach for:
| Tool | What it is simple for |
|---|---|
| Doodle | Finding a meeting time everyone can do |
| Google Forms / Typeform | Collecting structured answers and counting them |
| Skimle Ask | Collecting real opinions, in people's own words, and getting the themes back automatically |
What this is not
This is worth saying plainly, because it matters for how you should use it: Skimle Ask is not a replacement for proper qualitative interviews. If you are running a study that needs to hold up academically, navigating a sensitive or high-stakes topic, or trying to build real trust with a handful of people whose detailed perspective really matters, a skilled human interviewer is still the better tool for that job. Skimle Ask is the lightweight end of the spectrum: fast, low-friction, good for the questions where getting real answers from a lot of people quickly matters more than the depth you would get from a 45-minute conversation with each of them.
Think of it as the difference between a quick straw poll and a proper consultation. Both are useful. They are not the same tool, and using the lightweight one does not mean you have done rigorous primary research; it means you have got a fast, reliable read on what people think, which is often exactly what a casual situation calls for.
A practical workflow: launching your first Skimle Ask
1. Decide the one thing you actually want to know. "What do people think about returning to the office three days a week?" or "What's stopping customers from upgrading to the paid plan?" works better than a long list of topics. Skimle Ask defaults to a handful of core questions; resist the urge to ask everything.
2. Describe it in plain language and let Skimle draft the guide. You do not need to write survey questions from scratch, instead just write a sentence describing your goal, and Skimle's AI produces a short structured guide, typically three main questions plus a few quick multiple-choice ones if you want demographic or rating data alongside the open answers. From there you edit: reword anything that does not sound like you, add a topic you definitely want covered, delete one that does not matter and so on.
3. Set how deep each question should go. For each question, you can allow the AI zero, one, or several follow-ups. Leave follow-ups on for the question you actually care about, and turn them off for anything you only included for completeness. This is the dial that controls whether a respondent spends two minutes or ten.
4. Set a time limit and write a short, clear intro. A line or two explaining who is asking and why, plus an expected length ("this takes about 5 minutes"), noticeably improves both response rates and how candid people are. Decide whether responses should be anonymous; for anything even mildly sensitive, anonymous gets you more candid answers.
5. Share the link wherever your people actually are. Slack, email, a WhatsApp group, a QR code on a poster, an internal newsletter. There is nothing to install and nothing to log into. Respondents click, chat for a few minutes on their phone or laptop, and they are done.
For more detail on writing a guide that gets good answers rather than shallow ones, see how to write a good interview guide; the same principles apply whether the conversation is five minutes or fifty.
Analysing what comes back
This is the part that used to be the bottleneck, and it is the part Skimle Ask handles automatically.
As responses arrive, Skimle reads each one and starts grouping what people said into themes, the same thematic analysis approach used for full research projects, just running on a much smaller, lighter dataset. You do not just get a pile of transcripts to read one by one. You get a short list of themes, each showing how many people raised it and a few representative quotes, with every theme traceable back to the exact answer it came from, so if a finding surprises you, you can check it in seconds rather than take it on faith.
If you asked any multiple-choice questions alongside the open ones, you can slice the qualitative themes by them too, for instance seeing whether people in one department raised a different concern to another. The guide to working with metadata variables covers this in more depth, though for a quick poll you will usually just glance at the theme list and the quotes underneath it.
When you are ready to share what you found, you can export a short report, a few slides, or the full data as a spreadsheet, whichever fits the meeting you are walking into.
When to use Skimle Ask
A few situations where Skimle Ask fits naturally:
- Checking how your team feels about a proposal or document draft
- Getting a quick read on what a customer segment wants from the next release, beyond a star rating
- Running an informal pulse check across a community or alumni network
- Collecting input before an event or offsite, instead of an email thread nobody replies to fully
- A lightweight first pass research query before deciding whether a topic deserves a proper, in-depth research project
Frequently asked questions
Is Skimle Ask the same thing as Anthropic's interviewer tool?
No. They are built on the same underlying idea (an AI conducting a structured, adaptive conversation instead of a static survey), but Anthropic Interviewer is Anthropic's internal research tool, used to study how people use Claude. Skimle Ask is a productised general-purpose version anyone can use, for any topic, at any scale from ten people to thousands.
Do I need any research background to use this?
No. You describe what you want to find out in a sentence, and Skimle drafts the questions. You can launch a first poll without knowing anything about interview methodology. But at the same time being an expert in qualitative research allows you to conduct deeper surveys and analyse the results on another level.
How is this different from just adding an open-text box to a Google Form?
A text box collects whatever someone happens to type, with no follow-up if the answer is vague. Skimle Ask has an actual back-and-forth: if someone gives a one-line answer, the AI can ask a quick follow-up, the way a person would. You also do not have to read every response by hand afterward, Skimle groups them into themes automatically.
How long does it take to set one up?
A simple poll typically takes a few minutes to draft and configure. Respondents usually spend somewhere between two and ten minutes answering, depending on the time limit you set.
Is it free to try?
Yes, sending a Skimle Ask survey is free, and Skimle's free tier covers the analysis for smaller projects.
Ready to find out what people actually think? Try Skimle Ask for free and have your first set of answers, already organised into themes, within the hour.
Want the deeper version of this story? Read more on how Skimle Ask works end to end, or on when an AI interviewer is the right call versus a human one.
About the authors
Henri Schildt is a Professor of Strategy at Aalto University School of Business and co-founder of Skimle. He has published over a dozen peer-reviewed articles using qualitative methods, including work in Academy of Management Journal, Organisation Science, and Strategic Management Journal. Google Scholar profile
Olli Salo is a co-founder at Skimle and former Partner at McKinsey & Company where he spent 18 years helping clients understand their markets, develop winning strategies and improve their operating models. He has done over 1000 client interviews and published over 10 articles on McKinsey.com and beyond. LinkedIn profile