InsightGA4AttributionLead reporting
GA4 vs. last-click: why your best lead sources look broken in the default report
GA4's default attribution model buries the channels that actually start conversations. Here is how to diagnose what your reports are hiding before you cut a budget line.
Main takeaway
GA4's data-driven model and its summary reports are not the same thing. The default acquisition table shows last-click session data, not the model's output.
Best for
Service businesses running paid and organic traffic together
Time to apply
2 to 3 hours
Ship the first pass in this window
Recommended next step
Campaign ROI check
Pressure-test costs and conversion assumptions before your next budget call.
On this page
Why this matters
GA4's default Traffic acquisition report uses last-click attribution and ignores the data-driven model GA4 actually applies to conversions. The result: top-of-funnel channels that warm leads (organic search, referrals, display) get little or no credit in the report most teams read for budget decisions. To see the full picture, run a model comparison in Advertising > Attribution and adjust your lookback window to match your sales cycle.
The cost of skipping that diagnosis is real. Service businesses running annual budget reviews right now are looking at a full year of GA4 data, often their first complete year after migrating from Universal Analytics. Read that data at face value and the channels doing early-stage work get cut. Lead volume drops three to six months later, and the team has no clear reason why.
Last-click vs. data-driven attribution at a glance
| Aspect | Last-click attribution | Data-driven attribution |
|---|---|---|
| Where to find it in GA4 | Reports > Acquisition > Traffic acquisition | Advertising > Attribution > Model comparison |
| How credit is assigned | 100% to the final source before conversion | Distributed across touchpoints based on observed conversion patterns |
| Best for | Reading final-click channel volume | Identifying under-credited top-of-funnel channels |
| Limitation | Erases the role of awareness channels | Needs roughly 30+ conversions per event per month to stabilize |
| Default in GA4 | Yes, in summary reports and the acquisition table | Yes, but only inside the Advertising > Attribution section |
The attribution diagnosis framework
1) Identify which report you are actually reading (10 minutes)
Open GA4 and go to Reports > Acquisition > Traffic acquisition. The channel column here uses session-scoped last-click attribution by default. This is not the same as the data-driven model GA4 applies to conversions in the Advertising section. Write down which report you have been relying on for budget decisions. If it is the Traffic acquisition table, you have been reading last-click data regardless of what your GA4 attribution setting says.
2) Pull the model comparison report (20 minutes)
Go to Advertising > Attribution > Model comparison. Set the conversion event to your primary lead event, form submissions or calls, and compare last-click against data-driven side by side. Sort by the absolute difference in conversions assigned. Any channel where data-driven assigns materially more credit than last-click is a channel your budget decisions have been undervaluing.
3) Check your conversion event setup (15 minutes)
A model comparison is only as good as the events feeding it. Go to Admin > Events and confirm that your key lead events, contact form submissions, phone click events, booking completions, are marked as conversions. If only one event is marked, the model has a thin data set and will produce unreliable output. Service businesses typically need at least two to three distinct conversion events firing before data-driven attribution stabilizes.
4) Review the conversion path report (20 minutes)
Go to Advertising > Attribution > Conversion paths. Filter to your top conversion event. Look at the "paths" column for multi-step sequences. If most paths show only one touchpoint, either your tracking has gaps, users are converting on their first visit, or cross-device journeys are breaking the session chain. Note which channels appear most often as first touch and which appear most often as last touch. A channel that consistently appears first but rarely last is a candidate for under-credit in any last-click view.
5) Audit your lookback window (10 minutes)
GA4's default attribution lookback window is 30 days for most conversions. For service businesses with longer consideration cycles, a roofing job, a legal matter, a home renovation, 30 days may cut off early touchpoints entirely. Go to Admin > Attribution settings and check the window. If your average time from first visit to lead submission is longer than 30 days, extend the window to 60 or 90 days and note the date you made the change so you can compare before and after.
6) Cross-reference against a non-GA4 source (15 minutes)
GA4 will always have gaps because it runs in the browser and is subject to ad blockers, consent mode restrictions, and session fragmentation. Pull your CRM or form tool's raw submission log for the same date range. Compare total lead count to GA4's conversion count for the same event. A gap larger than 15 to 20 percent suggests tracking issues that will distort any attribution model, regardless of which one you use.
Scoring worksheet
8Total points
- Conversion events set up correctly and marked in GA40 to 2
- Model comparison report reviewed in the last 30 days0 to 1
- Lookback window matched to your actual sales cycle0 to 1
- Conversion path report showing multi-touch paths for at least some leads0 to 2
- GA4 lead count within 20 percent of CRM or form tool count0 to 2
This week action plan
- Day 1
Run steps 1 and 2. Write down which channels show the largest gap between last-click and data-driven.
- Day 2
Audit conversion events in GA4 Admin. Fix any that are missing or misfiring.
- Day 3
Pull the conversion path report and note first-touch vs. last-touch channel patterns.
- Day 4
Check and adjust the lookback window if needed. Log the change date.
- Day 5
Pull your CRM lead count for the last 90 days and compare to GA4 conversion volume. Document the gap.
Common mistakes
- Reading the Traffic acquisition table and assuming it reflects GA4's attribution model, when it reflects last-click session data.
- Running a model comparison with only one conversion event marked, which gives the algorithm too little data to produce reliable output.
- Adjusting channel budgets immediately after migration without waiting for the new model to accumulate enough conversion history, typically 30 or more conversions per event per month.
Frequently asked questions
Data-driven attribution requires a minimum number of conversions to function. Google has not published a hard threshold, but the model tends to fall back toward last-click behavior when conversion volume is low. A service business generating fewer than 30 to 50 conversions per month from a given event will likely see the model produce results close to last-click anyway. At that volume, the conversion path report and a manual CRM cross-reference will give you more useful signal than the model comparison.
Put this into practice
Turn this insight into a ranked homepage action list
Pressure-test costs and conversion assumptions before your next budget call.