Feedback pipeline: PostHog surveys for collection + AI triage job

Open
  • Is something wrong?
#104 2 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

Nobody has claimed this yet.

Assessment

Difficulty
5/5
Estimated time
Over a week
Newbie friendliness
25/100
Issue type
Feature
Clarity
Mostly clear
Activity status
Active
Tech stack
github, supabase, typescript
Domain
ai, analytics, backend, mobile, web-dev

Research direction

Start by locating the Expo app and web entry points to assess where PostHog SDKs would be wired; then identify how scheduled jobs and GitHub issue creation are handled. The issue specifies survey targeting and throttling, response triage, reply drafts, and reporter notifications, but does not name files or tests. Done means these collection and follow-up workflows are implemented with the stated safeguards.

Written by the indexing model from the issue text.

Description

priority-p1

Context

We want automated user feedback collection that isn't obnoxious, plus AI-assisted triage of what comes in — hands-off, fitting a solo maintainer.

Decision (2026-10-13)

Use PostHog Surveys for collection rather than a homegrown Supabase table:

  • No-code survey authoring from the dashboard (freeform text, NPS, ratings, multiple choice), popup or API/headless
  • Targeting by event / user property / feature flag (e.g. ask after 10th saved recipe — never on launch)
  • Built-in wait-between-surveys throttling so nobody gets nagged
  • Responses link to session replay + product analytics (see what the user was actually doing)
  • PostHog AI summarizes responses natively
  • Pricing: first 1,500 responses/mo free, then $0.10/response. Revisit if volume makes this expensive (5k responses ≈ $350/mo) — the Supabase-table fallback is documented below.

Work items

  1. SDK wiring: add PostHog to the Expo app (posthog-react-native) and web (posthog-js). Start with one survey: a single freeform question ("What's one thing we could do better?") targeted at engaged users, throttled.
  2. Triage job (scheduled): pull survey responses via PostHog API; classify (bug / feature request / confusion / praise) + sentiment/severity; cluster into themes with counts; open GitHub issues for confirmed bugs; emit a weekly digest (top themes, new bugs, urgent items).
  3. Reply drafts: AI-drafted responses to App Store reviews / support email — drafts only, never auto-send.
  4. Close the loop: notify reporters when their theme ships.

Collection rules

  • Never: popups on launch, repeat asks after a dismissal, blocking flows for feedback.
  • Native review prompts (SKStoreReviewController / Play in-app review) are separate; trigger post-success only.

Fallback

If survey volume ever makes PostHog pricing painful: replace collection with a feedback table in Supabase (user id, text, screen, app version, timestamp) + a small in-app form, keep the triage job pointed at the table instead.

Dominant language
TypeScript
Stars
0
Forks
0
Avg merge
8h 48m
Merged PRs (30d)
209

Getting set up

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from spudly/eciperay

All issues in spudly/eciperay

Similar issues

More TypeScript issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.