Historical backfill: what it changes for your KPIs | Guides
Beginner1 min

Historical backfill: what it changes for your KPIs

Understand the "Most recent leads are historical imports" banner. Why period-over-period comparisons are skewed after a backfill.

Why this banner appears

After a CSV import or CRM connection, Metrikia sometimes ingests several years of leads at once. The lead created_at stays the original source date (HubSpot, Pipedrive, etc.), not the Metrikia import date.

So a lead created in September 2025 in your CRM and imported in April 2026 appears as a September 2025 lead in Metrikia. Rolling "last 30 days" windows count these as new on the import day.

Why it is a problem

  • Period-over-period deltas spike artificially (+780% shows up after backfill)
  • The average CPL is diluted by leads without spend
  • Diana alerts trigger spuriously

How Metrikia handles this

  • The banner shows while more than 80% of leads have a created_at predating tenant creation
  • After 60 days of normal usage, the ratio drops and the banner disappears
  • The planned evolution adds an imported_at column to separate source creation from import

What to do meanwhile

  1. For historical analysis, use date filters on periods before the import
  2. For recent analysis, filter on the period after Metrikia activation
  3. Trust absolute values (closed deals, revenue) more than deltas during the post-backfill phase

If the banner persists

Beyond 60 days without fresh activity, the banner may indicate:

  • Your CRM no longer sends new leads (check webhooks)
  • The pixel is not installed (no organic leads tracked)
  • Ad tracking is broken (see dedicated guides)

To go further, check out our blog, the documentation or contact support.

Ready to take action?

Create your Metrikia account and apply this guide in minutes.