If signals are arriving late, three likely causes: provider delivery delays, integration backlog, or rate-limiting. Diagnose with the signal stream timestamps.
What 'latent' means
Signal latency = time between the underlying event (research activity, funding announcement, profile change) and the signal landing in turgo. Tight latency (under 1 hour) lets the agent act inside the buying window; loose latency (1+ days) is too late for most plays.
Provider delivery delays
Some providers batch and deliver once a day (Bombora has historically); some are near-real-time (G2). If your latency is consistently ~24 hours and your provider is daily, you've hit the provider's floor. Solution: combine with a near-real-time signal source for time-sensitive plays.
Integration backlog
If signals are batching at turgo's ingestion layer (the provider sent them in time but turgo is processing slowly), the integration log shows the backlog. The cause is typically a configuration issue or rate-limit hit. Check Settings → Signals → [Source] → Status.
Rate limiting
Provider APIs have rate limits. If you're querying frequently (e.g. real-time enrichment combined with signal pulls), you might hit limits and back off. The integration log shows rate-limit responses. Solution: increase the rate-limit budget with the provider, or shift to webhook-push instead of polling.
When latency is acceptable
Some plays don't need real-time. Job-change plays wait 30 days anyway, so 24-hour signal latency is irrelevant. Funding plays — fast is better but a day late isn't fatal. Latency matters most for intent surge plays where the buying window is hours, not days.
