Three tools, three different winners
A Kochi bakery chain asked which channel worked: Meta said Instagram drove 120 orders, Google Ads said 90, and GA4 credited most orders to direct traffic.
All three were partly right. Meta counted people who saw an ad and bought within a day; Google counted ad clicks; GA4 credited the last channel before purchase.
We stopped adding platform numbers together and judged the business on total orders vs total ad spend (MER), using each platform’s report only to compare ads within it.
Illustrative example based on a common pattern in Kerala e-commerce accounts. Numbers rounded.
What an attribution model is
Last click gives all the credit to the final click before conversion. Data-driven attribution looks at converting and non-converting paths in your account and shares credit across touches.
Ad platforms also use attribution windows: how many days after a click or view a conversion can still be credited. Longer windows and view-through credit make results look bigger.
Credit for a conversion = model (data-driven / last click) × window (days after click or view)
The models you’ll meet
Model or setting and what it does
| Model / setting | What it means |
|---|---|
| Data-driven (GA4, Google Ads) | Shares credit across touches based on your data |
| Last click | All credit to the final click before converting |
| Meta 7-day click, 1-day view | Counts clicks within 7 days and views within 1 day |
| Meta 1-day click | Stricter: only conversions within a day of a click |
How experienced marketers use Attribution model
1. Compare like with like
Use one model and window consistently when comparing periods or campaigns.
2. Judge prospecting fairly
Last click undervalues awareness campaigns; data-driven or blended MER gives a fairer view.
3. Reconcile with the CRM
Use CRM revenue as the source of truth and platform attribution to steer within each platform.
What good looks like
| Business type | Starting rule | Why |
|---|---|---|
| Google Ads default | Data-driven | Other rule-based models were removed in 2023. |
| Meta default | 7-day click, 1-day view | Compare with 1-day click to see view-through effect. |
| Source of truth | CRM or store revenue | Platforms each credit themselves. |
When the number lies
- Every platform claims the same sale
- Adding up platform conversions often gives more sales than you actually made.
- View-through credit
- 1-day view counts people who saw but didn’t click; useful, but easy to over-trust.
Mistakes to avoid
Adding Google and Meta conversions together
Double counting and an inflated ROAS.
Changing windows mid-month
Numbers jump without anything real changing.
Cutting awareness ads on last click alone
Lower-funnel campaigns slowly dry up.
Related terms
Questions people ask about an attribution model
What is an attribution model?
A rule for deciding which touchpoints get credit for a conversion, for example the last click or a data-driven split.
Which attribution model should I use?
Data-driven in Google Ads and GA4 for most accounts. Keep it consistent and check results against CRM revenue.
Why do Meta and Google report different conversions?
They use different models, windows and view-through rules, and each credits its own ads.
Is last click attribution still available?
Yes, in Google Ads and GA4, alongside data-driven. Other rule-based models like linear were retired.
Put it into practice
- SOP 313: See Campaign and UTM Results in GA4 →
- SOP 306: Audit a Google Ads Account and Write a Fix-First Plan →
Prefer to talk first? Book a free counselling call.