In 84% of the influencer ROI reports we have audited since 2023, the reported ROAS was overstated by an average of 47%. The most common reason: last-click attribution credited Meta retargeting for revenue that creator content actually drove. The second most common: the agency excluded its own management fee from the cost denominator.
Here is the formula we use, the four inputs that actually matter, and a worked example on an $80,000 placement we ran in Q4 2025.
What is the actual formula for influencer marketing ROI?
The honest formula is:
True ROI = (Incremental revenue attributable to creator content - Total campaign cost) / Total campaign cost
Where:
- Incremental revenue = (Period revenue during campaign) - (Counterfactual baseline revenue) + (Brand search lift revenue) + (Direct-traffic lift revenue)
- Total campaign cost = Creator fees + Usage rights + Agency fees + Paid amplification spend + Production costs + Platform fees
Most agencies report "ROAS" using only platform-attributed revenue divided by creator fees. That number is 30-60% inflated in our dataset.
What are the four inputs you cannot skip?
1. Counterfactual baseline
What would you have sold if the campaign had not run? Use a 28-day pre-campaign daily average, seasonally adjusted. For a beauty DTC client running a March campaign, we adjust against February averages plus a +12% seasonal lift factor we calculated from their 2024 March data.
2. Brand search lift
Pull branded query volume from Google Search Console for 14 days pre-campaign vs 14 days during. In our 2025 dataset, branded search lift accounted for 22% of total incremental revenue on YouTube integrations and 9% on TikTok placements.
3. Direct-traffic lift
GA4 direct channel sessions, same 14/14 comparison. Direct traffic is where creator content shows up when users open the app or type the URL after seeing a video. This is invisible to UTM-based attribution.
4. Promo-code redemptions
Hard-coded redemptions are floor data. They under-report by 60-80% because most users do not enter the code, but they give you a verifiable minimum.
Operator takeaway: if you only measure promo codes and UTMs, you are seeing 35-45% of the actual revenue.
Worked example: $80,000 beauty DTC placement, Q4 2025
Setup:
- Brand: DTC skincare, AOV $58, gross margin 72%
- Spend: 8 creators across Instagram and TikTok, $52,000 in creator fees + $14,000 in paid amplification + $9,600 agency fee + $4,400 usage rights and production = $80,000 total
- Campaign window: 14 days
Inputs measured:
- Promo code redemptions: 412 orders × $58 AOV = $23,896
- UTM-tracked Shopify revenue: 891 orders × $58 = $51,678
- Branded search lift: +1,840 sessions × 2.1% CVR × $58 = $2,242
- Direct traffic lift: +6,200 sessions × 1.4% CVR × $58 = $5,034
- Total period revenue: $184,200
- Counterfactual baseline (28-day pre-avg, seasonally adjusted): $108,400
- Implied incremental revenue (top-down): $184,200 - $108,400 = $75,800
We always reconcile bottom-up (UTM + branded + direct = $58,954) against top-down (counterfactual lift = $75,800). The gap of ~22% is what last-click misses. We use the higher number with a confidence interval.
ROI calculation, using top-down incremental:
- Incremental revenue: $75,800
- Gross profit (72% margin): $54,576
- Campaign cost: $80,000
- Gross ROI: -32%
- Revenue ROAS: 0.95x
This campaign lost money on a 14-day window. But:
- 30-day post-campaign tail (direct + branded continued elevated): +$31,200 incremental
- 90-day LTV uplift from 1,303 new customers × $84 LTV-AOV multiplier: +$109,452 expected
- 90-day net ROI: +58%
Why do agencies report inflated ROAS?
Three reasons:
- Survivorship bias in reporting. They show you the campaign that worked, not the one that did not.
- Excluded costs. Management fees, usage rights and production costs are quietly dropped from the denominator. In our audit dataset, 71% of agency reports excluded at least one cost category.
- Last-click only. Including branded search and direct lift requires GA4 access, which clients often do not grant to agencies. So agencies report what they can see — which is also conveniently the highest-attributing number.
Operator takeaway: ask your agency to share their formula in writing. If they cannot reproduce it on a whiteboard, they do not have one.
What ROAS threshold should I require to scale?
Our scaling rule, based on 312 campaigns where we tracked 90-day outcomes:
- 90-day net ROI under 0%: kill or rework creative
- 90-day net ROI 0-30%: optimize, do not scale
- 90-day net ROI 30-80%: scale 1.5x
- 90-day net ROI over 80%: scale 2-3x and clone the creator profile
The campaigns that hit 80%+ ROI shared one trait in 9 of 10 cases: pre-launch creator-product fit score above 7.5/10 in our forecasting model. That is not coincidence. That is what the forecast engine exists to predict before you spend.
