posthog/ai-plugin

modeling-revenue-metrics

Build reusable revenue models — MRR, ARR, gross revenue, new/expansion/contraction/churn, ARPU, LTV, and per-customer/per-account revenue — on either PostHog data-warehouse views (HogQL) or an external dbt project.

ソースを見る

使い方

  1. インストールコマンドをコピーし、自分の環境へ導入する前にソースを確認します。
  2. 必要なツール、認証情報、対応環境を原文で確認します。
  3. 入力と成果物を指定し、出力を資料と照合します。
npx skills add https://github.com/posthog/ai-plugin --skill modeling-revenue-metrics

開始前の確認

  • 導入にはモデル契約、API枠、ファイルへのアクセス許可は含まれません。
  • スター数や導入数は活動の指標であり、独立した品質・安全性試験ではありません。
リポジトリの原文

見出し、例、コード、表、リンク、参照画像を含む原文を表示しています。

Modeling revenue metrics

Turn payment/subscription data into durable revenue models. Read modeling-warehouse-foundations first for the view-vs-dbt decision, the view-* workflow, and convertCurrency(); this skill is the revenue-specific layer on top. Metric definitions live in `references/revenue-metric-definitions.md`; copy-paste recipes in `references/posthog/` and `references/dbt/`.

Step 1 — find where revenue lives

Revenue reaches PostHog two ways; both feed the same *managed `revenue_analytics_` views**:

  • A payment platform as a warehouse source — Stripe today (Chargebee/Polar/RevenueCat coming). Best when

the business runs on a billing platform. Connect via setting-up-a-data-warehouse-source.

  • Custom revenue events — you send events (e.g. purchase_completed) with a revenue property. Best when

there's no supported platform or you already track revenue in-product.

If neither exists yet, use suggesting-data-imports to recommend a source. In dbt, the equivalent is staging whichever billing tables landed in the warehouse.

Step 2 — model on the managed views, not raw tables

PostHog auto-generates a curated set of views per source. Do not re-derive revenue from raw Stripe tables — the managed views already handle deferred-revenue recognition, currency, and a stable schema.

Discover the exact names (they're prefixed by source, e.g. stripe.<prefix>.…, plus a cross-source revenue_analytics.all.…):

sql
SELECT table_name FROM system.information_schema.tables WHERE table_name ILIKE '%revenue_analytics%'
Managed viewGrainUse for
revenue_item (start here)1 / invoice line itemGross revenue, monthly recurring revenue, revenue by product/customer/period. Implements deferred revenue + currency.
mrr1 / (customer, subscription)Live snapshot of current MRR — not a time series.
customer1 / customerdim_customer: email, country, cohort, metadata.
subscription1 / subscriptionSubscription state for churn/expansion logic.
charge1 / chargeRaw charges; prefer revenue_item unless you specifically need charges.
product1 / productProduct dimension.

Key revenue_item columns: amount (already converted to the project base currency), currency (that base currency), original_amount / original_currency (as charged), is_recurring, customer_id, subscription_id, product_id, group_0_key…group_4_key (B2B account keys), timestamp.

Rules before you model (revenue gotchas)

  1. MRR is empty without a subscription config. For event-based revenue, MRR only populates when a

subscription property is configured. Empty MRR + populated gross revenue is expected behaviour, not a bug — say so instead of "fixing" it.

  1. The `mrr` managed view is a current snapshot, not history ("MRR at the current time"). For MRR _over

time, sum recurring `amount` per month from `revenueitem (see the recipe), or materialize a monthly snapshot of the mrr` view on a schedule.

  1. `amount` is already in base currency. Use it directly for reporting. Only call

convertCurrency(original_currency, 'XXX', original_amount, timestamp) when you need a different target currency, or when working from raw events.

  1. Link revenue to people via metadata. Person/group-level revenue needs

posthog_person_distinct_id metadata on the Stripe customer (or the person join). Without it, revenue is customer-level only.

  1. Don't build on the Revenue dashboard — it's being retired (~2026-06-30). Model against the

revenue_analytics_* views and the person/group revenue properties.

  1. Exclude test accounts. Confirm filter_test_accounts behaviour so QA/internal charges don't inflate

revenue.

Step 3 — build the model

PostHog: write the HogQL (alias every column), view-create, verify with view-get, then view-materialize the expensive monthly rollups (a daily sync_frequency is usually right for revenue). Recipes: `references/posthog/` — mrr_and_arr.sql, gross_revenue_by_month.sql, revenue_by_customer.sql.

dbt: stage the billing source → fct_revenue_item, fct_mrr, dim_customer marts with tests. Recipes: `references/dbt/`. Note dbt has no convertCurrency() — supply a rate seed.

Then register the model (references/governance.md in foundations): annotate columns and, if MRR/ARR is a headline number, propose it to the semantic layer.

File map

FileRead when
`references/revenue-metric-definitions.md`Precise definitions: MRR, ARR, gross, new/expansion/contraction/churn, ARPU, LTV.
`references/posthog/`HogQL view recipes on the managed views.
`references/dbt/`dbt staging + fct_*/dim_* marts + schema.yml tests.

Companions

modeling-warehouse-foundations (mechanics), setting-up-a-data-warehouse-source + suggesting-data-imports (get Stripe/revenue data in), modeling-dimension-tables (currency/plan dimensions), querying-posthog-data (HogQL + the semantic-layer metric check).

タスクに関連する項目

同じリポジトリから

関連する Skills

すべての Skills
posthog
公式

exploring-llm-costs

Investigate LLM spend in PostHog — total cost over time, cost by model, provider, user, trace, or custom dimension, token and cache-hit economics, and cost regressions. Use when the user asks "how much are we spending on LLMs?", "which model / user / feature is most expensive?", "why did cost spike?", wants to build a cost dashboard or alert, or pastes a trace URL and asks about its cost.

1時間の導入数
2
GitHub Stars
98
更新日
11月6日
posthog
公式

exploring-llm-traces

Debug and inspect LLM/AI agent traces using PostHog's MCP tools. Use when the user pastes a trace or session URL (e.g. /ai-observability/traces/ or /ai-observability/sessions/ ), asks to debug a trace, figure out what went wrong, check if an agent used a tool correctly, verify context/files were surfaced, inspect subagent behavior, investigate LLM decisions, or analyze token usage and costs. Also use when raw SQL/HogQL against events.properties.$aiinput / $aioutputchoices returns empty — message

1時間の導入数
2
GitHub Stars
98
更新日
11月6日
posthog
公式

instrument-integration

Add PostHog SDK integration to your application. Use when setting up PostHog for the first time or reviewing PRs that need PostHog initialization. Covers SDK installation, provider setup, and basic configuration for any framework.

1時間の導入数
2
GitHub Stars
98
更新日
11月6日
posthog
公式

investigating-error-issue

Investigates a single PostHog error tracking issue end-to-end. Use when the user provides an issue ID or pastes an issue URL (/error-tracking/ ) and wants to understand the error — who it affects, what triggers it, when it started, whether it correlates with a release, browser, OS, or feature flag, and what the next step should be. Pulls aggregated metrics, sample exception events, segment breakdowns, linked replays, and synthesizes a hypothesis-grade summary in one pass.

1時間の導入数
2
GitHub Stars
98
更新日
11月6日