Another week, another set of meaningful improvements to make your work with qualitative data faster, clearer, and more scalable. Here’s what’s new.

Explore Clusters in Chat with Labs 🎉

We’ve launched Labs - a focused experience for working with clusters inside chat.

Instead of long explanations, Labs puts clusters front and center. Answers are now cluster-only, with clear signal-style strength indicators that set expectations around precision. When you want to go deeper, simply click a cluster to open the Labs side panel for detailed exploration - without cluttering your chat stream.

The result: a cleaner, more trustworthy, and more scalable way to explore insights.

Hypothesis Validation Mode is Now Live 🎉

Chat has always been great for open-ended exploration. Now, it’s also built for structured validation.

With Hypothesis Validation Mode, you can test a specific assumption and get two clear evidence streams:

• Supporting evidence - statements that back up your hypothesis

• Contradicting evidence - statements that challenge or refine it

This gives teams a direct, evidence-based path to validate (or invalidate) assumptions using real qualitative data.

Copy-Paste Transcripts from NEXT 🎉

You can now copy full transcripts - including speaker names and timestamps - directly from the recording page.

Just select “Copy transcript” under Transcript options and paste clean, structured text into Google Docs or Word. Perfect for UXR workflows, internal reporting, and stakeholder sharing.

No more guesswork around who said what and when.

Smarter, Cleaner AI Cluster Creation 🎉

We’ve refined our AI cluster generation to improve both relevance and granularity.

By strengthening our prompts, we now:

• Enforce stricter alignment with the user’s question

• Prevent off-scope themes from creeping in

• Only merge clusters that are true duplicates

The result: more accurate, better-scoped, dimension-focused clusters that hold up under scrutiny.

Better Account-Based Search 🎉

You can now filter recordings by account and account metadata (industry, segment, etc.) directly from the recordings library.

We’ve also simplified account filtering across highlights. Instead of using recording.account.xxx, you can now use account.xxx everywhere - making search more consistent, especially for highlights coming from direct imports.

Finding the right data is now faster and more intuitive.

Hybrid Search for Chats: Keywords + Semantic 🎉

Chat search just got significantly more powerful.

We now combine semantic search with keyword matching in a single experience. This means you can:

• Ask natural-language questions for context-driven results

• Search for exact terms, names, or IDs with precision

• Mix both, without switching modes

Flexible when you need context. Exact when you need control.

Refined Wootric Mapping Is Live 🎉

We’ve rolled out a fully validated and more accurate Wootric field mapping.

This delivers cleaner, richer survey imports, including:

• Clear NPS/CES question prefixes in feedback text

• Normalized 5-star ratings

• Automatic score labels like “NPS: 9” or “CES: 4”

• Expanded tagging for edition, platform, user profile, industry, reseller, UI type, and more

The result: better-structured highlights, more context, and less manual cleanup.

More to come next week - as always, we’re building fast and listening closely.