Why Google Ads and GA4 Disagree on Conversions

A Google Ads vs GA4 conversion discrepancy under 15–20% is often normal—different attribution models, click vs session windows, and conversion counting settings explain most gaps. When Google Ads conversions and GA4 key events diverge beyond 20–25% for the same action and date range, run an elimination sequence before changing bids or client strategy.

Paid media managers live between two dashboards that rarely agree. Google Ads conversions optimize campaigns in real time; GA4 conversion data feeds funnels, LTV, and executive summaries. A conversion tracking mismatch feels urgent because both platforms claim authority. This guide applies a four-stage elimination framework (like traffic drop analysis), concrete magnitude thresholds, a decision tree for which number belongs in the client deck, and links to report methodology so stakeholders understand why one slide says 412 and another says 538.

What causes Google Ads and GA4 conversion numbers to disagree?

Root causes cluster into five buckets. Most accounts exhibit two or three simultaneously.

Fair competitor context: Triple Whale, Northbeam, and Rockerbox attempt multi-touch reconciliation outside Google—but they add cost and still depend on tag quality. For Google-only stacks, disciplined elimination beats buying a third number without fixing imports.

What is the elimination sequence for conversion discrepancies?

Work stages in order. Do not tune bids on Stage 1 noise.

Stage 1: Align definitions and date boundaries

Pass criteria: Same conversion action name, same timezone, same date range (calendar midnight boundaries), same property linked to Ads.

Checklist:

• GA4 key event matches the event imported or mirrored in Google Ads
• Google Ads conversion action includes website + imported GA4 only if intentionally blended—never mix lead forms from Ads UI with GA4 purchase events
• Ads account timezone matches GA4 property reporting timezone
• Compare conversion date in Ads vs event date in GA4—or align both to click date for paid-only views

Threshold: After alignment, remaining gap under 15% → document as expected variance; monitor monthly.

Fail: Gap above 20% with verified alignment → Stage 2.

Stage 2: Is the gap global or campaign-specific?

Segment Google Ads conversions and GA4 key events by campaign (and device if mobile-heavy).

Rule: If >70% of absolute gap sits in one campaign, fix that campaign before global tag rewiring.

Use a GA4-scoped Conversational Analytics thread: "Google Ads conversions by campaign last 28 days" in one chat, then a GA4-scoped thread: "GA4 key event purchase by session campaign, paid search only, same period." For side-by-side methodology in a client deck, use Report Builder.

Stage 3: Measurement integrity—tags, imports, and duplicates

Google Tag Manager audit triggers:

• Duplicate GA4 config tags firing twice (inflates GA4)
• Google Ads conversion tag missing consent triggers in EEA
• Conversion linker not on all domains in cross-domain setup
• gclid stripped by redirect—Ads clicks, GA4 attributes Organic

GA4 import health:

• Linked Google Ads account shows Import status: Active
• Auto-tagging enabled in Ads
• No conflicting Google Ads conversion actions for the same URL + event

Duplicate counting test: One test conversion on a staging URL excluded from production filters. Record whether Ads logs 1, GA4 logs 1. Multiple GA4 events per purchase (purchase + thank_you_page) with both marked key events can double GA4 vs Ads if Ads imports only one.

Threshold: Tag failure often produces >50% gaps or zero on one side—not gentle 18% drift. Zero GA4 with healthy Ads → GA4 event not firing or not marked key event. Zero Ads with healthy GA4 → tag or import break.

Cross-check site behavior during traffic anomalies with GA4 traffic drop vs Search Console—sudden session loss can depress GA4 conversions while Ads still records click-era conversions.

Stage 4: Attribution and reporting philosophy—which number for which decision?

After Stages 1–3, residual gaps often reflect attribution discrepancy, not broken tags.

Document the residual gap percentage after alignment. 10–20% residual with healthy tags is a normal google ads ga4 discrepancy band for prospecting campaigns with view-through influence.

What magnitude thresholds should trigger investigation?

Use these default agency thresholds; tune per client vertical.

Apply thresholds on volume-qualified data: minimum 30 conversions per side per period. Below 30, percentage gaps swing wildly—a 5 vs 8 event week is 60% mathematically but not operationally significant.

Week-over-week spike rule: If gap widens more than 10 percentage points week over week with stable spend, treat as incident even if absolute gap remains under 20%.

Which number should go in the client deck?

Use this decision tree:

Start: Same event, aligned dates, 30+ conversions each side
│
├─ Gap ≤15% → Client slide: Google Ads conversions
│              Footnote: "GA4 key event [name]: [X], within normal variance"
│
├─ Gap 15–20% → Client slide: Google Ads conversions (primary)
│                 Secondary bullet: GA4 figure + one-line attribution note
│
├─ Gap 20–35% after Stage 3 pass → Client slide: Both numbers side by side
│                 Narrative: "Platforms count differently; tags verified [date]"
│                 Do NOT average without CRM validation
│
├─ Gap >35% OR Stage 3 fail → Client slide: "Under measurement review"
│                 Show last known good week; withhold ROI claims
│
└─ CRM is source of truth contractually → Client slide: CRM conversions
                  Ads + GA4 in appendix only

Never present the higher number because it flatters performance. Never hide the lower number in appendix-only fine print if the deck headline cites ROI.

Report Builder methodology slides should state: conversion action name, counting method (one per click/session), attribution model, date basis, and residual gap band. That pattern matches the report methodology sibling guide—clients who read Report Builder HTML marketing reports expect explicit definitions on every conversion chart.

Pair ongoing monitoring with KPIs Tracker: set paired goals for Ads conversions and GA4 key events; Off track on both confirms systemic issue; Off track on one only points to attribution or import skew.

How do Google Ads conversions and GA4 key events count differently?

Example: Every purchase counted in GA4 but Ads imports one per click → GA4 routinely 15–40% higher on high-repeat purchase sites. Not a bug if documented.

Example: View-through display conversions enabled in Ads, GA4 key event web-only → Ads higher with gap 20–50% on display-heavy accounts.

What questions should you ask during a discrepancy review?

Borrow from the GA4 question library:

• "GA4 key event [name] by session google ads campaign, last 28 days"
• "Google Ads conversions by campaign, same period"
• "Sessions with gclid parameter present vs key event rate, paid search"
• "Landing page for Ads campaigns with >20% session drop and conversion drop"

For AI-influenced branded search lifting Ads conversions while GA4 attributes Organic, see tracking AI assistant traffic in GA4.

How often should you reconcile Ads and GA4?

Log each reconciliation: date, gap %, stage reached, resolution. Auditors and future you will need the paper trail.

Frequently asked questions