Influencer Marketing Conversion
Trace the $5.78 influencer ROI claim, distinguish discount-code attribution from incrementality, and assess follower size, disclosure, and Korean live-commerce figures.
Written and reviewed by Eric Han · Co-founder & CTO, reopt
Key takeaways
- The widely cited $5.78 appears to update an EMV-related figure from the previous year; the original was not verified. The 2019 edition reports EMV of 5.2 times spend but directly tracked ecommerce returns around 0.8 times, excluding later and non-click purchases.
- The mean ROIS of 17.85 for nano-influencers is based on discount-code attribution and is driven by extreme values. The overall median is 1.66.
- We found no independent peer-reviewed randomized experiment measuring sales. The grade A evidence found concerns TV viewing.
- Studies favoring nano versus large influencers use different outcomes and measures. Do not mix engagement with code-attributed revenue.
- Sponsored-post share increased after disclosure regulation. We found no quantified sales effect of disclosure.
What Is the $5.78 Claim?
- Branding campaigns8.0×
- Average EMV5.2×
- Direct response1.8×
- Tracked ecommerce returns0.8×
The $5.78 appears to be the following year's update in this series (2020 PDF inaccessible; summary only). The original states that 0.8 times excludes branding value, later purchases, and purchases without clicks. The $6.50 claim comes from self-reports by 125 marketers in 2015. Do not use these as inputs. TikTok Spark Ads claims of 142% and 43% circulating in blogs were also excluded because they could not be verified on TikTok's official pages.
Discount-Code Attribution: Beichert et al.
Beichert et al. (Journal of Marketing, 2024) analyze 2,808 codes, 1,698 influencers, and 1.88 million units sold in European D2C fashion. ROIS is (code revenue − cost) ÷ cost: revenue-based, not margin-based (observational, grade B). Costs are estimated by the authors using industry information rather than observed, making revenue per follower a safer input.
| Size | Mean ROIS | Revenue per follower |
|---|---|---|
| Nano, 1,219–8,496 followers | 17.85 | €0.265 |
| Micro | 5.98 | €0.080 |
| Macro, 49,845–1.4 million followers | 4.67 | €0.049 |
| Overall mean | 8.55 | €0.117 |
| Overall median | 1.66 | €0.030 |
Three field studies in the same paper report paid Instagram Story code/link revenue per follower of €0.001–0.008 for low-follower creators and €0.00006–0.0001 for high-follower creators. Approximate ratios calculated from means give revenue ÷ compensation of about 0.004–0.43, all below 1. Relative size comparisons may be robust while absolute returns depend on the sample. Management costs for many nano creators and long-term effects are excluded.
Code attribution can err in both directions
Codes miss people who buy without a code and can claim existing customers who search for one (author judgment). Without a no-ad control, incrementality is unknown. Compare total revenue across control regions or periods without influencer activity.
Incremental Evidence
| Evidence | Figure | Nature | Grade |
|---|---|---|---|
| Gong et al., JMR 2017 | Company Weibo tweets raised viewing 77%; influencer retweets added 33%. Authors' rough returns were 698% for tweets and 87% for tweets plus retweets, excluding carryover. Retweets alone yielded CNY 674 versus 1,000 cost, calculated | 98 randomized programs; TV viewing cannot be transferred to sales conversion | A |
| Meta partnership ads | Median incremental purchase cost −45% when partnership ads made up at least 30% of assets. 20 EMEA tests, January–June 2025, 92% confidence | Platform experiments; reviewed only via article | C (via press) |
| Yang et al., 2022 working paper, 2025 Marketing Science version unverified | A 1% increase in product-related engagement associated with +2.58% sales. Overall video engagement did not moderate sales lift | Observational identification; about 20,000 videos | B |
| IPA via Meta white paper | Short-term ROI index 99; long-term 151 | Modeled values; original unverified | C (via press) |
Yang et al.'s distinction between engagement and sales is particularly relevant in practice.
Follower Size: Separate the Outcomes
| Outcome | Direction | Evidence |
|---|---|---|
| Code-revenue ROI | Higher for smaller creators | Beichert et al., B |
| Engagement elasticity | A 1% spending increase yields 0.457% more engagement; positive size effect | Leung et al., 2022, B (summary) |
| Median engagement rate | Instagram nano 1.78%; macro/mega about 0.33–0.35% | HypeAuditor 2026 coverage, C (via press) |
Advertising Disclosure
Ershov and Mitchell (RAND, 2025) use difference-in-differences around tighter German regulation (12,000 creators, grade B). Sponsored-post share rose 4.6 percentage points, or 2–4.6 points under alternative classifiers, while likes fell more than 50% relative to the baseline mean. The authors identify engagement as an imperfect proxy for attention. This is not a sales effect. Another result reports a 0.19% subscriber decline for channels posting sponsored videos (Cheng, Zhang et al. press release).
Korean Fair Trade Commission detections totaled 108,709 in 2021–2025, but these are monitoring counts, not a violation rate. We found no Korean estimate of disclosure's sales effect. Legal disclosure duties apply regardless of effectiveness.
Korean Live Commerce
Most frequently quoted figures originate in estimates. Do not use eBEST's 2020 report figures of 5–8%, sourced only to “industry data, TWOAB,” the 20% Chinese wanghong claim, or one TMON broadcast's 2.2% (all D). Naver cumulative buyers ÷ cumulative views of about 0.9–1.0% (calculated, C), plus Naver/Kakao figures, appear in Korean Cases. We found no official value explicitly defined as a conversion rate.
Simulation Inputs
| Stage | Input | Evidence and conditions |
|---|---|---|
| Code revenue per follower | €0.00006–0.03 | Attributed. Overall mean €0.117 and nano mean €0.265 are tail-driven and useful only for checking an optimistic scenario |
| Attributed ROIS | Negative to 1.66 | Revenue-based; costs estimated by authors. Applying margin lowers returns further |
| Partnership-ad incremental cost change | 0 to −45% | Platform evidence; discount the −45% end |
Do not use $5.78, $6.50, 17.85, or live-commerce 5–8% / 20% as inputs.
Compensation and contribution margin are illustrative, except field-study preset followers and compensation, which use Beichert et al.'s means. We found no independent CPM/CPE benchmark; enter actual quotes.
Referral Experiments and Limits to Spread
Examine research on referred-customer value, incentive experiments, and viral diffusion, and why assuming K above 1 is risky.
Partnerships, Collaborations, and Sponsorships
Distinguish attribution and incrementality in affiliates, B2B partner win rates, podcast and newsletter sponsorships, and brand collaborations, with attention to denominators.