CHABOT

Analytics tools

4 minute read 818 words

DevRel teams measure two distinct things and need different tooling for each:

  • Product analytics. What developers do inside the product (signup, activation, retention).
  • Community analytics. What developers do in spaces around the product (GitHub, Discord, social, conferences).

The major tools for each category follow.

Product analytics

Mixpanel

  • Founded. 2009.
  • Positioning. Product analytics for tracking events, funnels, cohorts, retention.
  • Strengths. Mature funnel and retention analysis; strong cohort segmentation.
  • Used by. Many SaaS products including developer products for onboarding-funnel analysis.

Amplitude

  • Founded. 2012.
  • Positioning. Product analytics with stronger emphasis on experimentation and behavioural cohorts.
  • Strengths. Big-data scale; AI-assisted analysis; experimentation tooling.
  • Used by. Many large product-led-growth companies.

Heap

  • Founded. 2013.
  • Positioning. Auto-capture analytics: track everything and define events retroactively.
  • Strengths. No upfront instrumentation; flexible event definition.
  • Used by. Companies prioritising agility over schema discipline.

PostHog

  • Founded. 2020.
  • Open source. Yes: self-hostable.
  • Positioning. All-in-one product analytics, session replay, feature flags, A/B testing, surveys.
  • Strengths. Open source friendly to developer-product companies; strong default privacy posture; growing rapidly in 2024 to 2026 among developer-first companies.
  • Used by. Many YC-stage and developer-product startups.

June, Mixpanel-alternatives

  • June, Fathom (analytics), Plausible, Umami. Smaller alternatives; some privacy-focused.

Snowplow

  • Open-source event-data pipeline. Used by larger orgs to own their own event-collection layer rather than outsourcing.

Web analytics (for content)

  • Google Analytics 4 (GA4). Default for many docs sites; non-trivial migration from older Universal Analytics.
  • Plausible. Privacy-first, simple metrics.
  • Fathom. Similar.
  • Cloudflare Web Analytics. Free, privacy-respecting, simple.
  • Vercel Analytics, Netlify Analytics. Platform-bundled analytics.

Search and SEO tools

  • Ahrefs, Semrush, Moz. SEO tools for understanding how developers find content.
  • Google Search Console. Free search-performance and indexing data for docs sites.
  • Algolia DocSearch logs. What developers search inside your docs.

Community analytics

CNCF DevStats

  • Premise. Tracks all CNCF projects’ contribution activity (PRs, issues, comments, reviews) by contributor, country, organisation, over time.
  • Free, open-source. Dashboards at devstats.cncf.io.
  • Use. CNCF uses it to measure contribution activity across foundation projects.

LFX Community Data Platform

CHAOSS (Community Health Analytics for Open Source Software)

  • A Linux Foundation project. Defines metric standards for open-source community health.
  • Practitioner-led; provides metrics models that DevStats and other tools implement.

GrimoireLab

  • Open-source analytics platform for software-development data; underpins DevStats and Bitergia services.

Bitergia Analytics

  • Commercial. Analytics services built on GrimoireLab; used by many large open-source foundations.

Orbit (model) implementations

  • Orbit Model as a measurement framework can be implemented in any analytics stack. See Orbit Model.

Common Room, Savio

GitHub Octoverse / GitHub Insights

  • Octoverse is GitHub’s annual state-of-software report; built on GitHub’s internal analytics.
  • GitHub Insights (formerly part of Pull Panda; integrated into GitHub Enterprise) provides repo-level analytics for individual repositories.

Discord and Slack analytics

  • Native Discord analytics are limited; many teams supplement with custom bots writing to a warehouse.
  • Slack Analytics is limited at the Free tier and substantial at Enterprise Grid.
  • Common Room and Orbit-derived tools sit on top of these to provide cross-channel views.

Stack Overflow / Reddit / dev.to analytics

  • Stack Overflow tag-watching and search-data; modest in 2026 given platform decline.
  • Reddit’s official analytics for subreddit moderators; less useful for content discovery.
  • dev.to dashboard for authors and organisations.

Data warehouse and BI stacks

Teams that need consistent definitions across sources can pull data into a warehouse:

  • Warehouse. Snowflake, BigQuery, Databricks, Redshift, PostgreSQL.
  • Ingestion / ELT. Fivetran, Airbyte, Stitch; raw GitHub / Discord / Slack / Discourse / Postman / Salesforce API pulls.
  • Transformation. dbt for SQL modelling.
  • BI / visualisation. Looker, Mode, Hex, Metabase, Superset, Tableau, Power BI.

Pulling data into a warehouse gives the team control over definitions such as “active contributor” and “qualified developer lead,” which can otherwise differ between dashboards.

Trade-off: the team needs data-engineering capability, or access to someone who has it.

What to track

DevRel teams commonly track the following metrics. See ../04-metrics/ for detail.

  • Funnel: awareness → signup → first hello world → activation → retention.
  • Community: active users, contributor return rate, sentiment, NPS.
  • Content: views, completion rate, search queries, AI-assistant retrieval frequency.
  • Events: registration vs. attendance, post-event activation, share-rate.
  • Pipeline / revenue: DevRel-influenced revenue, DQL pipeline.

A common failure mode is to instrument everything possible and report on nothing. Pick the 10 to 20 metrics that connect cleanly to AAARRRP goals and refresh weekly. Everything else is for ad-hoc analysis only.

See also