Why "did the video do well" isn't the same as "did the campaign work"
Views and likes measure content performance, not commercial return. On TikTok Shop specifically, you have access to more direct commerce signals — GMV attributed to a creator, items sold, and average order value — which are far better proxies for whether a campaign actually moved product. This guide focuses on those.
The three numbers that matter most
- GMV attributed to the creator — total value of sales generated through their TikTok Shop links/tags over the campaign window
- Items sold — sales volume, useful for distinguishing "many small orders" from "few large orders"
- Average order value (AOV) — GMV divided by items sold; tells you whether the creator's audience buys in bulk or one-off
Benchmark: what "normal" looks like by tier
Among active sellers in our database, here's how average order value and sales volume vary by follower tier:
| Tier | Avg order value | Avg items sold (30d, active) |
|---|---|---|
| Nano (<10K) | $46.54 | 160.1 |
| Micro (10K–100K) | $43.83 | 696.4 |
| Mid (100K–500K) | $46.83 | 2,298.1 |
| Macro (500K+) | $31.87 | 4,179.4 |
Two things worth noting. First, average order value is fairly consistent across nano through mid tiers (roughly $44-47) — it doesn't scale up with follower count the way total GMV does. Second, macro creators show a lower average order value despite far higher volume, which is consistent with our tier expectations analysis: macro accounts often reach a broader, less purchase-intent-driven audience, converting more units at a lower price point rather than fewer units at a premium price.
A simple ROI framework
For a seeding/affiliate deal, ROI is straightforward: (GMV attributed − product cost of goods sold − affiliate commission paid) ÷ total cost (product + shipping + any flat fee). For a paid flat-fee deal, replace "affiliate commission" with the flat fee itself, and compare the resulting margin against your normal customer acquisition cost benchmark for other channels.
If you ran a multi-creator campaign, calculate this per creator, not just in aggregate — see our budget planning guide for tier-level cost context. Aggregating hides which specific creators actually drove the return.
What to do when a campaign underperforms
- Review the original screening record. Compare the creator's pre-campaign public signals with the criteria used for selection, without assuming those signals caused the result. See our red flags guide.
- Compare actual campaign inputs. Review price, offer, product availability, tracking setup and audience assumptions before attributing a result to the creator alone.
- Separate observed metrics from diagnosis. Strong views with low attributed GMV describe two campaign measurements; they do not by themselves prove an audience mismatch or explain causation. Use content review, tracking checks and your own customer data before drawing a conclusion.
Next: before your next campaign, build a vetted shortlist using the 7-step screening framework, and put terms in writing using our contract basics checklist.