Calculating Marketing ROI on Meta Ads After iOS 14
A working method for calculating honest Meta Ads ROI after iOS 14 — reconciling inflated platform ROAS against GA4 last-click and blended MER before it hits your ROI calculator.
Quick answer
Take Meta's reported attributed revenue, strip view-through conversions, discount the remaining click-attributed revenue by 20-40% (higher for considered-purchase and high-AOV brands), then sanity-check against GA4 last-click and blended MER. Use the reconciled number — not Meta Ads Manager's — as the revenue input in your marketing ROI formula.
Calculating Marketing ROI on Meta Ads After iOS 14
Computing honest Meta Ads ROI by reconciling Meta's overstated in-platform ROAS against GA4 last-click and blended MER before running the ROI formula.
Since Apple's App Tracking Transparency rollout in iOS 14.5, Meta's Events Manager has relied heavily on modeled conversions to fill in what it can no longer observe deterministically. The result: platform-reported ROAS typically overstates real revenue impact by 20-40%, with the largest gaps on considered-purchase categories.
Calculating true Meta Ads ROI means correcting for that overstatement before the number hits your ROI formula. The workflow is a three-way reconciliation: Meta's in-platform figure, GA4 last-click, and blended MER (total revenue ÷ total ad spend). Where the three disagree, you triangulate — and feed the reconciled attributed revenue into the calculator, not the platform number.
If you're running Meta at €80k+/month on a Shopify apparel or beauty store, the gap between Ads Manager ROAS and your bank account is probably the single largest measurement error in your reporting. Getting the ROI number right changes budget decisions.
This page walks through the reconciliation as an operational routine you can run weekly — the mechanism, the detection method, typical overstatement by vertical, and how the reconciled figure plugs into the marketing ROI formula.
Why Meta's reported ROAS overstates after iOS 14
iOS 14.5 stripped Meta of the deterministic pixel signal for opted-out iOS users — roughly 75-85% of iOS traffic in the EU. Meta filled the gap with statistical modeling and, critically, a much heavier reliance on view-through attribution.
Two mechanics inflate the number. First, view-through conversions credit Meta for purchases where the user saw an ad but never clicked — often purchases that would have happened anyway. Second, modeled conversions extrapolate from a shrinking pool of observed data, and the model is trained by Meta to be generous. The deeper mechanics are covered in the companion page on quantifying Meta's post-iOS14 conversion overstatement.
The view-through problem in one line
A 1-day view-through window means anyone who saw your ad on Instagram and bought within 24 hours — even from a direct email click — counts as a Meta conversion. Turn view-through off in reporting before you do anything else.
How to detect your overstatement gap
Run three numbers side by side for the same 28-day window: Meta's reported attributed revenue (7-day click, 1-day view), GA4 last-click revenue attributed to Facebook / Instagram paid, and blended MER (total store revenue ÷ total ad spend across all channels).
Meta will almost always be the highest. GA4 last-click will be the lowest — it undercounts because it ignores discovery. Blended MER is the honest ceiling: if you paused Meta entirely, how much revenue would actually disappear? For a full walkthrough see reconciling Meta ROAS with GA4 last-click for DTC stores.
The gap between Meta-reported and GA4 last-click is your raw overstatement signal. Divide GA4 by Meta — a ratio below 0.6 means Meta is claiming credit for revenue GA4 can't corroborate. That delta is where your discount lives.
Typical overstatement by vertical and AOV
Meta-reported vs GA4 last-click revenue gap, typical ranges by vertical (post-iOS 14, 7-day click / 1-day view)
| Vertical | Typical AOV | Meta overstatement | Recommended discount |
|---|---|---|---|
| Impulse beauty / cosmetics | €35-60 | 15-25% | 20% |
| Fast fashion / apparel | €60-120 | 20-30% | 25% |
| Premium apparel | €120-250 | 25-35% | 30% |
| Home & furniture | €200-600 | 30-40% | 35% |
| Considered electronics | €300-1200 | 35-50% | 40% |
| Jewellery / watches | €400-2000 | 40-55% | 45% |
The pattern is consistent: the longer the purchase consideration window, the more Meta over-claims. High-AOV brands see the worst gap because view-through and modeled conversions have more time to spuriously align with organic demand — the mechanics are unpacked in why high-AOV considered-purchase brands see the biggest Meta overstatement.
The reconciliation math
Start with Meta's reported attributed revenue. Strip view-through conversions — typically 15-25% of the reported total on a 1-day view window. Apply your vertical's discount to the remainder. Then compare to GA4 last-click × 1.4 (a rough uplift to account for GA4's own undercounting of paid social discovery). If the two numbers land within 15% of each other, you've triangulated cleanly.
If they diverge more than that, blended MER is the tiebreaker — you compare your current MER to your MER during a Meta-paused period or a geo holdout. This is where a geo holdout test to back out Meta incrementality becomes worth the effort, and where blended MER as the tiebreaker earns its name.
Feeding the reconciled number into the ROI calculator
The marketing ROI calculator takes attributed revenue, ad spend, gross margin, and other costs. Post-reconciliation, the only field that changes vs the naive workflow is attributed revenue — you use your reconciled figure instead of Meta's screenshot. Everything downstream (payback, contribution margin, ROAS floor for bid strategy) inherits the correction.
One extra step matters: if your CAPI implementation is weak, Meta's model is working with even worse input signal and overstates more. Audit CAPI event match quality (aim for 8.0+) before you trust any of these numbers — see how CAPI implementation quality changes your Meta ROI math. Then push the reconciled ROI back into bidding as covered in feeding reconciled Meta ROI back into bid strategy.
Meta Ads ROI after iOS 14 — FAQ
For DTC e-commerce, typically 20-40%. Impulse-purchase categories with low AOV see the smallest gap (15-25%); high-AOV considered-purchase categories like jewellery or premium furniture routinely show 40-55% overstatement versus GA4 last-click.
Neither on its own. Meta overstates by modeling and view-through; GA4 last-click understates because it ignores discovery-led paid social. Reconcile the two, cross-check with blended MER, and use the reconciled figure in your ROI formula.
Blended MER (Marketing Efficiency Ratio) is total store revenue ÷ total ad spend across all channels. It's platform-agnostic and immune to attribution modeling, so it acts as the honest ceiling when Meta and GA4 disagree on channel-level ROAS.
Yes. View-through credits Meta for purchases where the user saw an ad but never clicked — often revenue that would have converted through email or organic anyway. Switch reporting to 7-day click, 0-day view, or subtract the view-through line explicitly.
Weak CAPI (event match quality below 6) means Meta's model has poor deterministic signal and modeled conversions are less reliable — overstatement worsens. Fixing CAPI first can shrink the reconciliation gap by 5-15 percentage points before you touch anything else.
Weekly if Meta is more than 30% of your paid mix, monthly otherwise. Meta's modeling drift and seasonal shopping behaviour both change the overstatement ratio, so a discount you calibrated in Q1 will be stale by Q4.
Either GA4 tracking is broken (check cross-domain, consent mode, and server-side events) or Meta is over-claiming aggressively. Run a geo holdout — pause Meta in one region for 3-4 weeks and measure the revenue drop against a matched control region.
Yes — anything reported in Meta Ads Manager runs through the same modeling and view-through logic. Reels in particular tends to over-claim because impression volume is high and users often see an ad without clicking.
Partly. Meta and GA4 both expose the raw numbers via API; the discount factor and blended MER cross-check require judgement calls that change with seasonality and CAPI health. A weekly dashboard with a manual override on the discount ratio is the pragmatic setup.
Most brands find their pre-iOS14 Meta ROAS was roughly 15-25% lower than the platform reports today — closer to today's reconciled figure. Backfilling a pre-iOS14 baseline from GA4 historical import gives you an honest before/after and confirms the reconciliation is landing in the right zone.
Track CAC, channels, and funnel conversion in one place
Metricuno connects ad spend, funnel events, and revenue so you can see CAC by channel, cohort, and campaign — without stitching together five tools.