Track interview completion
- Interview details now show Complete when the participant has answered every guide question and applicable follow-up, alongside the closing message.
- Use Complete only in the interview filters to review finished interviews.
Follow the participant's language
- Video interviews continue in the language of a participant’s answer when that language is enabled for the study.
- Read-aloud follows each question’s language, and mixed-language interviews can be translated back into their starting language.
Video interviews
- Set a study’s Interview type to Video interview and participants answer on camera.
Study Builder v2
- Build, edit, and preview your study in one redesigned workspace.
Agent v2
- The agent now replies noticeably faster during interviews, so participants spend less time waiting between answers.
Introducing Interview guide
- Choose between two styles: Questions the agent asks word for word, or Goals the agent explores in its own words.
- Mix both in one guide, switch an item’s type at any time, and add Follow-ups instructions to steer how the agent digs deeper.
- Interview depth now sets how many follow-ups each item gets, from none on Express to three or four on Deep, with an estimated interview length beside it.
Pick prize-draw winners
- Draw your winners straight from the Reward tab. Each person gets one entry no matter how many interviews they completed, and every winner links back to their interview.
Dark mode
- The dashboard now has a dark theme. Switch between Light, Dark, and System from your account menu at the bottom of the sidebar.
Pillow has a new look
- Every screen has been redesigned for a calmer, more readable workspace.
Branded study links and custom domains
- Your studies now live on a branded
<your-org>.pillow.sosubdomain, so your links show your organization’s name instead of a generic address. - Add a Custom domain in Settings → Domains to serve studies from your own subdomain, such as
interview.acme.com.
Suggested studies and follow-ups
- Find Suggested Studies tailored to your product and past findings, right in the new-study page.
- Each study’s executive summary proposes Follow-up studies grounded in its findings.
Revamped Insights page
- The Insights page has been revamped, starting with a brand-new Executive Summary of your study.
- Issues now surface feature requests alongside bugs and UX problems, clustered across every interview.
- Every question cluster in the Overview now includes its own sentiment analysis.
Faster, more natural interviews
- Interviews now respond faster, keeping conversations flowing without long pauses.
- Replies feel more natural and conversational.
Connect AI tools with the Pillow MCP
- Connect Claude, ChatGPT, Cursor, and other AI tools directly to Pillow and explore your study results without exporting transcripts.
- Ask questions across your studies, search interviews, and pull recurring issues and patterns straight into your workspace.
- Access is read-only and scoped to one organization: sign in with your Pillow account, no API keys to manage.
- Open the new MCP tab in the dashboard to connect. See Connect AI tools with MCP.
Synthetic study testing
- Run synthetic participants against a test study to see how your agent handles several realistic interviews before launch.
- Watch multiple test interviews unfold side by side.
- Choose participant locales and inspect participant profiles to check how the study performs across different audiences.
Filter and search your studies
- A new filter bar on the studies list lets you narrow by status, creation date, interview count, reply rate, and creator.
- Active filters appear as editable chips, and search jumps straight to a study by name.
- See Browse and filter studies.
Research Chat is now Ask Pillow
- The chat panel for questioning your interview data is now called Ask Pillow.
- A first-time hint points you to it above your study’s analysis tabs.
- See Ask Pillow about your data.
Custom agent roles
- Customize the role shown under your agent’s name in the interview chat header.
- Auto-translate the role for supported participant languages when setting up multilingual studies.
Reach participants in their language
- Add an outreach message for every supported language in one click with the new all-locales button.
- The interview experience now adapts to each participant’s own language automatically.
- See Run studies in multiple languages.
Smarter Study Builder
- The Study Builder can now search the web to ground your study in real context.
- It proactively suggests goals and shows its reasoning in a live action feed as it works.
- See Create a study.
Interview grades
- Each interview now gets a letter grade from A+ to D so you can spot the most valuable responses at a glance.
- Filter the interview list by grade to jump straight to your best interviews.
Participant rewards and identifier collection
- Offer participants a reward for completing an interview.
- Collect an email, phone number, or custom identifier from each participant, then export the list as a CSV.
- See Offer rewards and collect identifiers.
Target SDK audiences
- Launch studies automatically to specific cohorts of SDK users, no shared link needed.
- Build cohorts by combining activity rules with device, location, and custom property filters.
- See Target SDK audiences.
Mobile SDKs for iOS and Android
- Native SDKs let you embed Pillow studies directly in your iOS and Android apps.
- Identify users and set custom properties so studies reach the right audience segments.
- Control when and how studies appear with presentation options like skipping already-seen studies or forcing a fresh session.
- Track study lifecycle events: know when a study is presented, completed, skipped, or fails to load.
- Install via Swift Package Manager on iOS or Maven Central on Android.
Issue tracking
- Pillow now automatically detects bugs and issues mentioned during interviews and clusters them across interviews.
- Browse issue clusters in the new Issues tab with priority levels, verbatim participant quotes, and interview counts.
Add images to study goals
- Attach up to four images per goal to give the Study Builder and Research Agent multimodal context. Share screenshots, mockups, or reference visuals alongside your text.
Voice messages in interviews
- Participants can now send voice messages during interviews instead of typing.
- Voice messages are automatically transcribed so you can read them alongside text responses.
- Play back original audio recordings from the interview transcript.
Export Research Chat to PDF and Notion
- Export any Research Chat thread to PDF.
- Export any Research Chat thread to Notion.
Edit your study setup from the chat
- Ask the Study Builder to update agent settings, outreach messages, persona attributes, and interview depth directly in the interview.
- See inline previews of every change the builder makes, and click to review the full details.
- Drag-and-drop goals editor with inline editing, custom goal creation, and suggested choices above the chat input.
Study setup quality checks
- A background quality judge now evaluates your goals and notes, surfacing warnings when something looks off.
QA distribution charts in Research Chat
- Research Chat answers can now include QA distribution charts as inline evidence.
Share Research Chat with your team
- Share Research Chat threads and individual messages via a link.
- Choose between organization-only or public visibility, with optional expiration.
Verbosity control for Research Chat
- Control response length in Research Chat with Concise and Detailed modes.
- Concise mode produces tight summaries while still performing thorough analysis.
- Choose your preferred style once, and it persists across your chat session.
Translate interviews on demand
- Translate interview transcripts into your preferred language from the interview view.
- Read translated text inline while keeping the original messages available.
Filter Research Chat by metadata
- Narrow Research Chat analysis by metadata like locale, country, device, and custom fields.
- Discover available metadata values before filtering to focus answers on the right audience segment.
Notes to guide AI interviews
- The Study Builder agent can now take notes to guide AI interviews.
- Use notes to give extra context the AI should keep in mind during interviews.
Better interview view
- We redesigned interviews to make them easier to read.
Research Chat history fixed
- Reopening an older Research Chat now shows the thread as expected.
Filter interviews by language
- You can now filter interviews by locale.
PII improvements
- Sensitive-data detection was updated to reduce false positives.
Set default outreach locale
- You can now choose a default locale for outreach messages.
Interview summaries
- Each interview now gets an AI-generated summary with a headline and key bullet points.
- Hover over any interview to preview its summary at a glance.
Research Chat
- Ask questions about your study data and get AI-generated answers in a new chat interface.
- Insights are grounded in your actual interviews with source references.
- A small-sample banner warns you when data is limited so you can interpret results with the right context.
PII automatic detection
- Personally identifiable information (PII) is now automatically detected and can be redacted from interview transcripts and insights.
- Redacted content appears as styled labels instead of raw markers, making transcripts easier to read.
Interview depth control
- A new depth slider lets you choose between four interview styles: Express, Quick, Balanced, and Deep.
- Each level adjusts how many questions and follow-ups the agent asks, from fast broad coverage to thorough exploration.
New study builder and guided setup
- Introduced a new Study Builder with AI-powered goal generation.
- Guided setup makes it faster to turn objectives into actionable studies.
- Improved launch flow clarity so teams can go live with confidence.
Richer insights and smoother navigation
- Insights now include fuller interview context and study metrics.
Test & launch experience refined
- Preview chats directly from the test & launch page before sharing.
- Study tables and stats are clearer at a glance.
- Copy updates make settings and workflows easier to understand.
Study duplication and integration improvements
- Duplicate studies to reuse settings and move faster.
- Integration links are clearer with live mode guidance built in.
- Interview selection and Safari form behavior are more reliable.