Postback vs Meta conversions — which to trust

Two attribution sources: Meta-reported conversions (pixel + CAPI deduplication, Meta's algorithm) vs Postback-reported conversions (server-to-server from tracker, no browser dependency). They often disagree by 5-20%. Meta tends to over-attribute (claims credit broadly); postback is closer to actual. For ROAS optimization: use Meta's view (matches Meta's bidding signals). For business reporting: use postback (truer attribution). Document the discrepancy + use consistent source per use case.

Who is this for

Affiliate marketers + e-commerce mediabuyers running tracker-based funnels. Anyone whose Wevion Ads Manager shows N conversions while postback dashboard shows M conversions for the same campaign.

The two attribution sources

Source 1: Meta-reported conversions

How Meta tracks:

• Browser pixel fires on conversion event

• Plus CAPI (server-side) fires same event

• Meta's algorithm attributes the conversion to an ad based on attribution window (7-day click + 1-day view default)

• Deduplicates pixel + CAPI via event_id

• Reports back to advertiser

Visible in: Meta Ads Manager + Wevion's per-platform Meta column.

Source 2: Postback-reported conversions

How postback tracks:

• Tracker server logs conversion (Keitaro, Voluum, etc.)

• Fires HTTP POST to Wevion's postback URL with click_id

• Wevion matches click_id → ad campaign attribution

• Counts as conversion attributed to that ad

Visible in: Wevion's postback dashboard + per-campaign postback column.

See com-109 postback overview.

Why they disagree

Net result: Meta often over-attributes by 5-20%. Postback is closer to actually happened.

When to trust which

Trust Meta for: bidding + optimization

Meta's algorithm makes bidding decisions based on its own attribution. If you tell Meta "optimize for purchases" + Meta sees 100 purchases, it optimizes based on those 100.

If you only feed Meta postback-reported (say 70 purchases), Meta optimizes based on 70 — and may under-perform because it has less data.

So: use Meta's CAPI / pixel data for optimization signals. Don't try to reconcile Meta's bidding-side view with postback.

Trust postback for: business reporting + reconciliation

For "did we actually make money?": postback is closer to truth.

• Revenue calculation: postback total

• Profit calculation: postback total minus costs

• Client reporting: postback (defensible / auditable)

• LTV analysis: postback (matches actual buyers)

Document the discrepancy

Don't pretend they agree. Document for each campaign:

• Meta-reported conversions: X

• Postback-reported conversions: Y

• Discrepancy: Z%

If discrepancy > 25%: investigate (something likely wrong with one source).

Reconciliation workflow

Monthly reconciliation:

1. Pull Wevion's Meta-reported total

2. Pull Wevion's postback-reported total

3. Calculate discrepancy

4. Look for patterns: certain campaigns / dates / placements with bigger gaps?

5. Document acceptable threshold (10-20% typical)

6. If beyond threshold: investigate (pixel issue, postback issue, click_id flow)

Per-campaign attribution choice

Wevion can display either source per campaign:

• Default view: Meta-reported (most common workflow)

• Toggle: Show postback alongside

• Both visible in detail drawer per campaign

For Cross-Channel Analytics: typically Meta-reported (consistency with other platforms' algorithm-based attribution).

For exports / client reports: pick consistently (don't mix).

Wevion's role in reconciliation

Wevion does NOT decide which source is "right". Wevion presents both + tracks the discrepancy.

You make the call per use case.

For affiliate marketers: postback often primary source (Meta data less critical when most traffic isn't directly attributed to Meta).

For direct e-commerce: Meta primary (pixel/CAPI reliable for own-site conversions).

Common patterns by business model

Direct e-commerce (Shopify / WooCommerce)

• Primary source: Meta pixel + CAPI

• Secondary: own analytics (GA4, Shopify Order Attribution)

• Postback: optional if you also use a tracker

Affiliate marketing (Keitaro funnels)

• Primary source: postback (truer to actual)

• Secondary: tracker's own attribution

• Meta-reported: for optimization signals only, not business truth

Mixed (own site + affiliate)

• Two streams of data:

  • Direct: Meta CAPI primary

  • Affiliate: postback primary

• Separate reports per stream

• Don't blend; clarity matters

Disagreement size as signal

• 0-5%: noise; ignore

• 5-15%: normal; both functioning

• 15-30%: investigate (pixel or postback issue)

• > 30%: likely a setup problem (pixel broken, postback misconfigured, attribution model mismatch)

What you'll see in Wevion

Per-campaign detail:

• Meta-reported column

• Postback column (if postback configured)

• Discrepancy % indicator

Cross-Channel Analytics:

• Default: Meta-reported (for consistency)

• Toggle: switch to postback for affiliate-heavy workspaces

Common questions

• "My Meta says 100 conversions, postback says 75. Which is right?": both, for different purposes. Real conversions probably closer to 75 (postback). Meta sees 100 because of view-through + cross-device.

• "Should I disable Meta CAPI and just use postback?": NO. Meta needs CAPI for bidding optimization. Use both.

• "My discrepancy is 50% — what's wrong?": investigate. Likely: click_id not flowing through to tracker, or pixel only firing on certain conversion paths.

• "Can I get one unified number?": no — they're fundamentally different attribution methods. Pick one consistently per use case.

Best practices

Pick primary source per business case

Don't switch mid-month. Pick + document + stick with it.

Run both for redundancy

Even if you primarily trust one: run both. Lets you detect issues with either source (if one drops to 0, you know something broke).

Educate stakeholders

Clients / executives often confused by discrepancy. Explain: Meta's view = what Meta credits itself with; postback view = what your tracker confirms. Use the right one for the conversation.

Reconcile quarterly minimum

Periodic check prevents drift. If Meta + postback drift apart over time: investigate.

Related

• Postback overview — postback mechanics

• Postback vs pixel — which to use — companion comparison

• Campaign metrics decoded — interpreting metrics