Send Research Insights to Slack: Real-Time Customer Interview Notifications via Webhooks

The 60-second answer

Koji has no native Slack app and no outbound webhooks, so nothing inside Koji posts to Slack on its own. Two things do work today. The quickest: publish the study report and paste its public link into the channel. It stays live and updates as responses land, and anyone can open it without a Koji account. The automated one: create a Slack incoming webhook, then have your own backend, or Zapier, n8n or Make, poll the Koji read endpoint and post a message when an interview comes back completed with its analysis. A per-minute poll puts the quote, theme, quality score and transcript link in your channel about a minute after the interview ends.

If you have ever wished your team got pinged in Slack the moment a customer says "I'd pay double for that feature" — this is the workflow.

Why pipe research into Slack?

Research loses 80% of its impact when insights sit in a dashboard nobody opens. A 2024 Forrester study found that organizations with real-time research distribution channels were 3.5x more likely to act on customer feedback within a sprint, vs. teams reading quarterly reports.

With Koji feeding a Slack channel you get:

• Near-real-time visibility: every completed interview posted within a minute of finishing
• Cross-functional reach — engineers, designers, support, and execs all see the same raw quote
• Faster product decisions — a single quote in #product-channel can shortcut a 30-minute meeting
• Customer obsession at scale — your CEO can lurk on #voice-of-customer and stay close to users without scheduling a single call

Koji's read endpoint returns the finished interview; Slack's incoming webhooks accept structured JSON; a 30-line poller or a scheduled no-code automation glues the two together. We will show both.

What your poller can detect today

Koji emits no events, so the poller decides what counts as news. The states worth watching for, all visible on the read endpoint:

• status is completed: the interview finished, and the transcript and structured answers are ready
• analysis is populated: themes, the summary and the quality score are ready too
• the quality score is 3 or above, the same threshold Koji bills at
• a report exists for the study, so you can link straight to it

Themes across interviews are not pushed anywhere. To alert on one, keep a running count in your own store as each interview lands, or read the study report and let its clustering do the work.

For most teams the high-signal moment is an interview arriving with its analysis. Set up one channel for that.

Setup: Slack → Koji in 3 steps

Step 1 — Create the Slack incoming webhook

1. In Slack, go to Apps → search for Incoming Webhooks → click Add to Slack.
2. Pick the channel where research insights should post. We recommend #voice-of-customer for general research, #cs-alerts for churn signals, and #product-feedback for feature requests.
3. Slack generates a unique URL like https://hooks.slack.com/services/T0000/B0000/XXXXXXXX. Copy it.

Step 2 — Decide on payload shape

Slack's incoming webhooks accept either a plain {"text": "..."} payload or rich Block Kit JSON. The interview you read back from Koji looks roughly like:

{
  "status": "completed",
  "study_id": "abc123",
  "study_name": "B2B onboarding research",
  "respondent_id": "anon_xyz",
  "duration_seconds": 720,
  "quality_score": 4,
  "summary": "Participant struggled with onboarding step 3...",
  "top_quote": "I almost gave up at the import screen.",
  "themes": ["onboarding friction", "data import"],
  "transcript_url": "https://www.koji.so/p/abc123/r/anon_xyz"
}

Since the shapes differ, you need a tiny middleware that reformats Koji's JSON into Slack's format.

Step 3 — Connect with one of three integration patterns

Option A — No-code (5 minutes)

Use a tool like Zapier, Make, or n8n with this recipe: run on a schedule every few minutes, call GET https://www.koji.so/api/v1/interviews/{id} for each interview you are waiting on, filter to the ones that are completed with an analysis, transform with a "format JSON" step, then post to Slack with the incoming webhook URL. There is no Koji trigger app to subscribe to. Map the Koji top_quote field into the Slack text field with formatting like:

New interview completed for {{study_name}}
Quote: "{{top_quote}}"
Themes: {{themes}}
Quality: {{quality_score}}/5
Full transcript: {{transcript_url}}

Option B — Lightweight serverless (15 minutes)

Deploy a small scheduled worker on Vercel, Cloudflare Workers, or AWS Lambda. The worker polls the Koji read endpoint, reformats each finished interview into Block Kit, and pushes it to Slack. A complete TypeScript example lives in our Webhook Setup guide.

Option C — Direct via the Headless API (advanced)

If you are already using the Headless API, have your existing backend poll Koji for finished interviews and post to Slack alongside the work it already does (write to your data warehouse, update CRM, ping PagerDuty for negative sentiment, etc.). This is the route most enterprise customers go.

Step 4: Give the poller a key and a list

There is no webhook page in Koji to register a URL with. Instead, create an API key with the interview:read permission and keep it server side, then give your poller a list of interview ids to watch. Store each interview_id when you create or send an interview, and drop it from the list once you have posted it to Slack. See API Key Permissions and Rate Limits and CORS: a key allows 60 requests a minute by default.

Run the poller once by hand against an interview that has already finished. Within a few seconds you should see a message appear in Slack.

Five Slack research workflows worth setting up

1. Interview firehose

Channel: #voice-of-customer. Trigger: the poller sees an interview completed with quality_score 3 or above. Posts the top quote, themes, and transcript link. Keeps the whole team close to customers.

2. Churn-signal alerting

Channel: #cs-alerts. Trigger: the poller matches the interview's themes against cancel|leaving|switching|too expensive. Tags your CS lead. Pulls churn signals out of noisy data immediately.

3. Feature-request capture

Channel: #product-feedback. Trigger: the poller matches themes against wish|need|missing|wanted. Auto-creates Linear tickets via the Linear Slack app. Closes the loop from interview to backlog.

4. Daily research digest

Channel: #research-daily. Posts a 5pm summary every day: number of interviews completed, top 3 themes, top 3 quotes, quality score average. Ideal for execs.

5. Stakeholder mentions

Channel: project-specific (e.g., #proj-checkout-redesign). Filters interviews by study_id and posts to the matching project channel. Keeps domain stakeholders in the loop without spamming everyone.

Why a live channel beats the "export CSV every Monday" workflow

Platforms like Koji eliminate the "research bottleneck" entirely. Insights flow into the same channels where decisions already happen.

Filtering noise: only Slack the signal

Two controls keep your channels useful:

1. Quality gate: every conversation gets a 1 to 5 quality score, and only 3 and above is billed. Have the poller skip anything below 3 so low-effort or abandoned interviews never reach Slack. See How the Quality Gate Works.
2. Theme thresholds: there is no theme event to configure, so keep a running count in your own store and post only once 3 or more respondents have raised the same theme. Prevents single-data-point overreaction.

For sensitive employee research running in Anonymous Mode there is nothing to strip: the lead form is off, so the interview carries no name or email and your Slack message can only ever be theme-level.

Combining with Koji's 6 structured question types

When your study uses Koji's structured questions (open_ended, scale, single_choice, multiple_choice, ranking, yes_no), the interview you read back includes the exact structured answer for every question. This means your Slack message can say:

"Sarah scored NPS as 9/10. Top reason for the 9: 'The new dashboard saves me 2 hours a week.'"

…instead of just an unstructured quote. Single, parseable, ready to ingest into your analytics warehouse alongside the conversation.

Security best practices

• Keep the Koji API key server side. It carries read access to your studies, so it never belongs in browser code or in a client-side automation step.
• Store the Slack webhook URL as a secret in your environment (not in code). Anyone with the URL can post to your channel.
• Use a per-study webhook URL if you have sensitive studies. Slack lets you create unlimited incoming webhooks; isolating them per channel limits blast radius.
• Rotate keys quarterly. Issue a fresh per-study API key and retire the old one.
• Apply Koji rate limits — see Rate Limits and CORS. High-volume studies (1,000+ interviews/day) should batch events through a queue.

What about Microsoft Teams, Discord, or Mattermost?

The same pattern works. All three accept incoming webhooks. Replace the Slack URL with their equivalents. The Koji side is identical: your poller simply points at a different destination URL.

Plan availability and pricing

The API and the MCP connector are available on every plan, including a free account, and outbound webhooks are not available yet. Interviews start as low as €1 per qualified interview, and €3 per qualified voice interview. Reading interviews is not metered: you can poll as often as the rate limit allows without additional cost. See the plan comparison guide for the full feature matrix.

Related Resources

• Webhook Setup: why there are no outbound webhooks yet, with polling and embed-event code samples
• Research Automation: building near-real-time research pipelines
• Headless API Overview — running Koji programmatically
• Structured Questions in AI Interviews — the 6 question types that produce parseable data
• How to Automate User Research — building 24/7 research pipelines
• MCP Authentication & Security: keeping your Koji keys and tokens safe
• Real-Time Research Insights — the Koji insights dashboard

Further reading on the blog

• Best AI Tools for UX Research in 2026: The Complete Buyer's Guide — The UX research AI tool landscape has exploded. This guide maps the best AI tools for every phase of the research workflow in 2026 — from pl
• The UX Researcher's Guide to Scaling Research with AI (2026) — Demand for user research has never been higher — but researcher headcount hasn't kept pace. Here's how UX researchers are using AI to scale
• Why AI Interviewers Are the Future of Customer Research — AI interviewers are transforming how product teams conduct customer research, enabling conversations at scale without sacrificing depth or q

<!-- further-reading:blog -->