Survicate

Survicate connects to RudderStack. You already track what your users do in RudderStack. Now that behavioral and attribute data can flow into Survicate! Survey the right people at the right moment and read the results by segment. What you can do: → Sync user attributes, company attributes, and events from RudderStack with Survicate → Target surveys by specific RudderStack events → Coming soon: add RudderStack context into Research Hub and filtered survey analysis data Connect in one click 🤫

Stakeholders likely won't read your research preamble. So start with a visual. Nikki Anderson tested Research Hub's chart builder. She could: → Build charts from the underlying research. Sentiment, feedback count, chart type. Based on analyzed sources only. → Download charts in PNG and PDF. → Drop charts into reports with /chart. If a stakeholder can reach the decision faster from your report, the report gets used. That's what Research Hub is built around. First project on us if you want to try it on your own data → link in comments.

so, turns out your respondents are not nearly as flaky as you think we looked at 8,392 surveys, and the median completion rate came out at 81.9%. which means, once someone actually starts your survey, on average they get through about 82% of it. i used to tell my clients to keep their surveys super short. Now thinking, so much of survey design is built on fear! we tiptoe around our own audience like they're one question away from rage-quitting. don't get me wrong, short surveys still hold up especially well (2-3 questions land around 87%), but the longer ones do better than the doom suggests :D but my new advice: add that one extra question. Because data shows that length is hardly the culprit, format is. it swings results 17 points on its own, with link surveys finishing way ahead of the in-chat ones. what matters is ask well, put it in the right place, and trust people to show up. it seems, they usually do. full completion-rate report here if you want the format + industry breakdowns: https://lnkd.in/ek6G4mA9 what's your completion rate sitting at? or are you tracking it at all/if not why?

[Now live] Google Meet is a feedback source in Research Hub. 🥳 Because some of your sharpest customer feedback may never make it into analysis. It gets said out loud on a call, but then sits in a transcript nobody opens again. Now those calls land in your feedback stack automatically. Along surveys. Along other feedback sources. Giving you nothing but a clear picutre. Connect Google Meet to Research Hub and your meeting transcripts sync straight in, ready to be analyzed next to all your feedback. What you can do: → Import past transcripts from the last 7, 14, or 30 days, or pick specific meetings → Auto-sync new transcripts as they happen, with optional filtering by meeting title → Read an AI summary of every meeting, plus the full transcript → Pull it all into a Research Project and let the AI surface themes across sources and calls

how long should my survey take? ok so we actually measured it. across 7,677 surveys, the median completion time is 108 seconds. under two minutes 🤯 which is kind of reassuring? most surveys people fill out are quick. you're not making anyone sit through a saga. but "108 seconds" on its own is a bit of a lie, because completion time depends massively on context. a 3-question smiley survey is ~24 seconds. add a few open-text questions and you can 10x that without noticing. industries swing wildly too (e-learning clocks ~51 seconds, educational institutions closer to 6 minutes). so the number i want you to stop chasing is some universal "ideal" completion time. it doesn't exist. the only honest benchmark is your length + your channel + your audience. what i'd actually do? find the shortest version that still answers your question, then respect that you're spending someone's minutes either way. we broke the whole thing down by length, question type, channel, and industry. how long are your surveys running? be honest 👀

What's a good survey response rate? We analyzed 4,332 surveys from 460 companies across all of 2025. The median response rate came out at 9.98%. So if you're hovering around 10%, you're right in line with most companies. Here's a rough way to place yourself: → 0-4%: room to grow (common for exit widgets and broad, unsegmented sends) → 5-10%: good, where most companies sit → 11-22%: great, usually high-engagement formats or loyal segments → 22%+: excellent, the top 25% of performers A few things that move the number more than you'd expect: Where you ask. Mobile surveys, living inside iOS and Android apps, hit an 18.69% median. Website widgets sit closer to 8%. Same audience, very different result. Who you ask. B2C surveys see a 12.85% median, nearly double B2B at 8.18%. What industry you're in. Fintech sees a 20.11% median while Media barely goes above 5%. Even SaaS lands below the overall benchmark at 7.74%. Once people start, they tend to finish. The median completion rate is 75%, in around 42 seconds. Don't measure yourself against one global average. Compare to your format, your audience, and above all your own trend over time.

Welcome to Survicate, Yousician! Plenty of teams collect feedback. Far fewer turn it into something they act on continuously. Yousician, the app that's helped millions learn to play guitar, piano, ukulele, and bass, learn to sing, or simply get in tune with GuitarTuna (loved internally at Survicate too! 😉), switched to Survicate. Here's what Yousician were after in terms of feedback: → Native in-app surveys. Ask for feedback right inside the app, instead of bouncing users out to a browser. → Feedback wired into their stack. Responses connect to the tools they already use, so the right follow-up hits the right note for the right user at the right moment. → Continuous research. Always listening to what users actually need, from churn signals to checkout friction. All making feedback a habit, a full cycle, instead of an off-key project. That's the part we love most. Welcome aboard, Yousician. Here's to staying in tune with your users. 🎶

A 3-question survey takes about 24 seconds to finish. A 50-question one…closer to 41 minutes. We looked at 7,677 surveys from 1,044 companies, and survey length is by far the biggest thing shaping how long people spend. Roughly what each length costs your respondents: → 1-3 questions: ~24 seconds → 4-10 questions: ~1.5 minutes → 11-20 questions: ~7 minutes → 21-30 questions: ~18 minutes → 31-50 questions: ~34 minutes → 51+ questions: ~41 minutes The overall median across everything is just 108 seconds, but that's mostly because short surveys dominate: around 75% of all surveys are 10 questions or fewer. The practical takeaway is to know the time tax before you hit send. Every question you add is a small withdrawal from someone's day, and the jump from "quick favor" to "real commitment" happens faster than most people expect, somewhere around the 10-question mark. So when you're building, ask what the survey is worth to the person answering it, not just to you, and match the length to that. Full completion-time benchmark, with breakdowns by channel, question type, and industry in the comments.

What's a good survey completion rate? If your respondents answer about 80% of your survey questions on average, you're right where most companies are. Across 8,392 surveys from 1,087 companies in 2025, the median completion rate landed at 81.9%. Half of all surveys finished at or above that. A rough way to place yourself: → Below 68%: worth a look (bottom quarter of surveys) → 68-81%: solid, the typical range → 82-91%: strong → 92%+: excellent, the top 25% What actually moves the number? Format, more than anything else: Page/link surveys finish highest at 86.8%. Intercom surveys sit lowest at 69.4%. That's a 17-point swing from delivery alone. Shorter is smoother. 2-3 question surveys complete at 86.8%, with the steepest drop landing in the 4-6 question range. Business model barely matters this time. B2B (82.4%) and B2C (81.4%) finish at almost the same rate. That last point is the interesting one. In our response-rate report, B2C pulled nearly double the responses of B2B. But once people start, both finish at about the same rate. Getting someone in the door and getting them to the finish line are two different games. One more thing: completion isn't all-or-nothing. Survicate saves answers as respondents go, so even a half-finished survey still gives you usable insight. Eylül Nowakowska Beyazıt broke it all down by format, industry, and length in the full report. Three days, three stats! Day Two

"The only repositories I've seen really work is if you hire a librarian. Someone in charge of maintaining it, sharing it, making sure it actually works." That's Nikki Anderson on our webinar on surviving the AI sprint. But it's not just Nikki saying that. The Nielsen Norman Group asked 400+ UX and ResearchOps pros about their repositories. Only 9% described them as thriving. 11% described them as stagnant or forgotten. Almost 20% said adoption was poor. A repository on its own is a storage place. Without someone curating, surfacing, and re-sharing, it becomes an archive of last quarter's research that no one opens. That's the constraint we wanted out of the way when we built Research Hub. The AI Research Assistant answers questions across all your past research, sourced to the exact quote and participant, so "is this in there already?" stops being a someone-has-to-go-find-it problem. Themes and sentiment are tagged at the sentence level by AI, so categorization isn't a manual job. The researcher's time should go on the thinking. Not on keeping the system alive.