Back to integrations
Integration

Website embed — Insight feedback widget

Paste a short script to launch branded Insight sessions from live pages. Visitors explain friction in context; sessions use the same blueprint fields, optional voice, and editable analysis hub as link and CRM studies — conversational website feedback, not a heatmap, session replay, or star-only poll.

High-friction URLs — pricing, checkout, docs, onboarding — are where visitors feel stuck. Analytics show where they leave; they rarely explain why. The website embed puts an Insight agent on those pages so people can talk through the friction while the context is still on screen.

You publish one study, copy the widget snippet, and paste it on target routes. Pulse session shape keeps completion high in-page. Blueprint fields still capture structured scores and categories; the agent asks follow-ups when answers are vague. Results land in the same analysis hub as every other distribution channel.

Embed is not a replacement for Hotjar-style heatmaps or FullStory-style replays. Those tools show behavior. KogniFeed captures explained reasons from your owned audience — with free sessions to prove the page is worth instrumenting before you scale.

When to use this

  • A specific URL (pricing, checkout success, docs, help) shows drop-off or support tickets you cannot explain from analytics alone.
  • You want in-context feedback without sending visitors away to a long form or panel screener.
  • Pulse-length interviews fit the page better than a deep-dive CRM wave — or you will pair embed pulse with a later CRM deep dive.
  • You need the same editable metrics and quotes product already trusts from link or CRM studies.

Example use cases

  • Pricing and plan-page friction

    Embed on pricing to learn which plan comparisons confuse buyers — blueprint fields for intent, blockers, and competitor named.

  • Checkout or order-success pulse

    Capture post-purchase CSAT and fulfillment themes while the order is fresh, then compare periods after ops changes.

  • Docs and onboarding help

    Ask readers what blocked setup on the doc they just used; feed themes into product and documentation owners via the hub.

Setup steps

  1. 1

    Publish the Insight study

    Use a website-feedback template or your own brief. Prefer pulse session shape for in-page attention spans; configure blueprint fields for the scores and categories you need in charts.

  2. 2

    Open Share → embed

    Copy the widget snippet for your published study. Confirm branding, welcome copy, privacy link, and whether voice or text mode is available.

  3. 3

    Paste on target pages

    Add the script to pricing, checkout success, docs, or help articles. Test on staging before production so layout and consent copy look right.

  4. 4

    Optionally gate or target

    Show the widget on specific routes or after key actions. Avoid spraying every page — high-friction URLs convert better and keep themes interpretable.

  5. 5

    Review in the analysis hub

    Filter by date and compare periods after UX changes. Edit auto-metric labels when themes need tightening; export or share quotes with product owners.

  6. 6

    Iterate session shape and placement

    If completion dips, shorten welcome copy or move the widget later in the flow. Pair embed pulse with CRM deep dives when you need richer quotes from known contacts.

FAQ

Does embed replace Hotjar or FullStory?

No. Those tools show what users do. KogniFeed embed captures why — via AI-moderated conversation on the same pages.

Can participants use voice in the widget?

Yes, when voice mode is enabled in Share settings and the browser supports microphone access.

Is developer time required?

Only to paste and deploy the snippet. No custom interview UI to build — the Insight agent, blueprint, and analysis hub are already configured in admin.

Will this feel like a form-builder popup survey?

No. Visitors enter a branded conversational Insight session with adaptive follow-ups and structured blueprint fields — not a static checkbox poll skinned as chat.