Episode 11: Using Skimle for customer insights — and why metadata variables change the game
A listener got in touch to ask whether Skimle is only for interview-heavy academic or consulting work, or whether it can also handle customer research. The short answer is yes, very much so, and this episode is our longer answer.
We walk through how product teams, market researchers, and customer success functions are using Skimle to make sense of the qualitative signal that sits inside customer interviews, user feedback, support tickets, call logs, NPS verbatims, and anything else that comes in text form. The platform treats all of these as documents in a project, so you can upload a mix of sources and analyse them together — which is often exactly what you want when you are trying to build a picture of a customer segment rather than summarise a single data type.
The feature we spend most time on is metadata variables. When you upload your documents, you can attach structured metadata to each one: customer segment, product tier, region, interview date, renewal status, or any other dimension relevant to your analysis. Once the AI has coded your data and built the theme structure, those metadata variables become filters. You can ask: what do enterprise customers say about onboarding, versus what SMB customers say? How did the response to a product change compare across cohorts before and after a major release? Which themes are consistent across all segments, and which are specific to one group?
This kind of segmented analysis is where synthesising user research at scale becomes genuinely useful rather than just fast. Without metadata filtering, you get one big pool of themes. With it, you get a structured comparison across the dimensions that actually drive decisions.
We also discuss practical setups for customer research teams: Zoom and Teams call transcripts from discovery or renewal conversations, combining interview data with survey verbatims, and feeding in support ticket exports alongside structured interviews. The always-on customer research model — where Skimle Ask runs continuously to collect qualitative signal between research sprints — also gets a mention.
If you are doing competitive or win-loss research, competitive intelligence from customer interviews covers how metadata variables help there too. And for teams thinking about how to present findings once the analysis is done, presenting qualitative findings to executives is worth a read.
The customer and market researchers use case and product managers use case pages on our site show more about how different teams use this in practice.
Episode 12: Agentic AI — what it means and how Skimle is already there
For most people, AI still means a chat box: you type something, you get a reply. That prompt-response loop has been enormously useful, but it is not how the most capable AI systems work in 2026. This episode is about what comes next, and why it matters for qualitative research.
Agentic AI means giving a language model access to tools — the ability to read files, run searches, update records, trigger actions — so it can work toward a goal across multiple steps rather than just responding to a single question. Instead of asking "what do my customers say about pricing?" and getting an answer based on whatever the model can recall from context, you ask the question and the AI agent goes and looks. It queries your structured data. It checks themes. It pulls relevant quotes. It compares across segments. It tells you what it found and why.
Skimle's new AI chat works exactly like this. When you ask it a question about your analysed data, it does not guess or hallucinate from a summary — it uses tools to access the actual coding, retrieve quotes, and answer based on what is in your project. You can ask it to reorganise categories, move an insight to a different theme, or flag responses that mention a specific competitor. It acts more like a smart research assistant than a text generator.
This matters for trust. The transparency problem with using general-purpose AI for qualitative analysis has always been that you cannot see how it got from data to conclusion. Two-way transparency — being able to trace every claim back to source — is a core design principle in Skimle, and the agentic chat extends this rather than undermining it. You see what tools were called and what data was retrieved.
We also talk about why standard RAG approaches fall short for structured qualitative analysis, and why the tool-use model is a better fit for research corpora where the relationships between themes, insights, and quotes matter more than raw similarity search.
Shortly, these same capabilities will be available directly to external AI agents via an MCP (Model Context Protocol) interface — meaning tools like Claude or other agents your organisation uses will be able to query and interact with your Skimle data directly. If that sentence sounded like gibberish, do not worry: we explain what it means and why it is worth caring about in terms that do not require a computer science background. The short version is that your research data becomes something your broader AI toolchain can actually use, rather than a silo you visit separately.
For the broader context on where AI and software are heading, death of SaaS or a renaissance of better software is a companion read. And if you want to understand how LLMs actually process and analyse data under the hood, how ChatGPT actually works is a good grounding piece.
About Skimle and Skimlecast
Read more about Skimlecast and watch Episode 1 here! You can watch all our episodes on our Spotify or YouTube channels. If you have any comments, feedback, topics you would want us to cover or other things you want to share, please connect with us through our form or via email through olli@skimle.com.
If you are interested in using Skimle, check out how Skimle works. You can also try Skimle for free and see how AI-assisted qualitative analysis handles everything from academic interview data to customer feedback at scale.
Meet the cast
Henri Schildt is a Professor of Strategy at Aalto University School of Business and co-founder of Skimle. He has published more than a dozen peer-reviewed articles using qualitative methods, including work in Academy of Management Journal, Organisation Science, and Strategic Management Journal. His research focuses on organisational strategy, innovation, and qualitative methodology. Google Scholar profile
Olli Salo is a former Partner at McKinsey & Company where he spent 18 years helping clients understand the markets and themselves, 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