Skip to main content
Beton
posthog_marketplace posthog_marketplace_mvp_v1

Platform / marketplace

Two-sided marketplace where active listings are the key metric. Positive signals chain from account connection through storefront activation to listing publication.

Value metric

Listings, storefronts, connected accounts, integrations

Success event

listing_published

entity_type: account

Scale

  • 240 accounts
  • 5 users/account (mean)
  • 8 sessions/user (mean)
  • 30 days of history

Research metrics proxied

  • — Active listing growth
  • — Seller/buyer activation rate
  • — Marketplace liquidity ratio

Signal paths

Positive paths end in listing_published. Negative paths do not. Every generated event_id belonging to a path is recorded in ground_truth.json.

Positive signals (2)

connected_storefront_listing ×70
account_connected → storefront_activated → listing_published
cohorts: high_intent, medium_intent, power_user
integration_listing_publish ×45
integration_enabled → listing_published
cohorts: high_intent, medium_intent

Negative signals (2)

connected_storefront_stall ×55
account_connected → storefront_activated
cohorts: medium_intent, low_intent, lurker
integration_backtracking ×35
integration_enabled → account_connected
cohorts: low_intent, noisy_bot_like

Generate this dataset

Config file: configs/posthog_marketplace_mvp.yaml

Quickstart 
# Dockerized Postgres (recommended for inspection)
docker compose up -d

uv run dryfit \
  -c configs/posthog_marketplace_mvp.yaml \
  --dsn postgresql://dryfit_writer:dryfit_writer@127.0.0.1:54329/dryfit \
  --print-summary

# Or local Postgres
./scripts/generate-local -c configs/posthog_marketplace_mvp.yaml --print-summary

Full setup instructions are in the repo's README — including local Postgres, Grafana inspection, and dataset restore.

Noise parameters

DryFit injects realistic noise on top of the generated signal paths. These probabilities are per-event. Noise never touches rows referenced by ground_truth.json — your scoring logic can trust the truth file is exact.

missing event probability
5.0%
duplicate event probability
2.0%
out of order probability
3.0%
null property probability
3.0%
anonymous actor probability
1.0%
weird property probability
1.0%

Other scenarios

Combined coverage (all models)

→ upgrade_clicked

Union of event types across all business models

Contact / record-based SaaS

→ contact_created

Contacts, leads, subscribers, accounts managed

Credits / token-based

→ credits_purchased

Credits consumed, tokens used, compute units

Event-volume SaaS

→ custom_event_tracked

Events tracked, data points ingested, log lines
Browse all scenarios →

Benchmark your detector against Platform / marketplace

Clone the repo, run the config, check your agent's output against ground_truth.json.

View on GitHub