Guides · For shops and brands
Which KPIs really matter in influencer marketing?
The KPIs that matter depend on the goal. For sales: click-through rate (CTR), conversion rate, cost per order (CPO) and ROAS, meaning revenue divided by all campaign costs, ideally after returns. For awareness: reach, views and watch time per post. Engagement rate and follower count say little about revenue. Work everything out per creator too, or the average hides who sells. A pure last-click model does not credit creators for purchases made via search days after the click.
Last updated:
With wyrld Boost you put your budget behind creators who feature your products, starting today, and pay for results.
Nine metrics and how to calculate them
| Metric | Formula | Tells you |
|---|---|---|
| Reach | accounts reached, from the creator’s insights | how many accounts saw the content |
| Views and watch time | plays and time watched, from the creator’s insights | whether people actually watch the video |
| Engagement rate | interactions ÷ reach × 100 | whether the content resonates with the audience |
| Click-through rate (CTR) | clicks ÷ reach × 100 | whether the content makes people want the product |
| Conversion rate | orders ÷ clicks × 100 | whether the shop, price and offer win people over |
| CPO (cost per order) | cost ÷ orders | what one order costs you |
| ROAS (return on ad spend) | revenue ÷ cost | how much revenue each euro of cost brings back |
| New customer share | first orders ÷ all orders × 100 | whether the campaign brings in new customers |
| Revenue after returns | revenue minus returns and cancellations | what is actually left of the revenue |
Cost means everything the campaign costs you: fees, commissions, products sent out, shipping and, if you are liable for it, the artists’ social security levy (Künstlersozialabgabe) on the fees. If you only count the fee, ROAS looks better than it is.
Tools and agencies calculate engagement rate in different ways: by followers, by reach or by views. The same goes for click-through rate: according to its help pages, Google Ads divides clicks by impressions, meaning the number of times an ad is shown, not by accounts reached. Only compare figures that were calculated the same way. And take care with Google Analytics: there, engagement rate means the share of sessions on your website that last longer than 10 seconds, include a key event or have at least two page views. It has nothing to do with likes and comments.
Worked example: one campaign from post to ROAS
- Say you pay three creators €1,000 in fees between them and send out products worth €200. Your cost: €1,200. All amounts are net of VAT, and shipping and levies are left out to keep it simple. These figures are made up and are not benchmarks.
- The posts reach 80,000 accounts and get 3,200 interactions. Engagement rate: 3,200 ÷ 80,000 × 100 = 4%.
- The links get 1,600 clicks between them. Click-through rate: 1,600 ÷ 80,000 × 100 = 2%.
- They lead to 40 orders, averaging €75 net. Conversion rate: 40 ÷ 1,600 × 100 = 2.5%. Revenue: €3,000. CPO: €1,200 ÷ 40 = €30.
- ROAS before returns: €3,000 ÷ €1,200 = 2.5. In Google Ads you enter target ROAS as a percentage, so 2.5 corresponds to 250% there.
- Assume 30% of the revenue comes back as returns. That leaves €2,100, and ROAS after returns is 1.75.
- Now check your margin. Say half of net revenue is left after the cost of goods. Of €2,100 that is €1,050, less than the €1,200 cost. At this margin the campaign only pays for itself from a ROAS of 2 after returns.
- Repeat purchases may still make the campaign pay off. That depends on how many new customers you won. Say 26 of the 40 orders were first orders: that is a new customer share of 65%.
Why the last click can understate creators
An example: someone taps the link in a story on Monday, looks at the product, searches for it on Google on Thursday and orders. In a last-click model the last channel before the purchase gets all the credit, in this case Google search. The creator gets no credit in the report, even though their link brought the first visit.
According to its help pages, Google Analytics now offers only three models in its attribution reports: data-driven attribution and two last-click models. Data-driven attribution uses your account’s data to share the credit for a purchase across the clicks before it, so the story link from the example can get a share too. Google Analytics marks the purchase event as a key event by default and looks back 90 days for it, and you can also choose 30 or 60 days.
Web analytics only sees what reaches your website. If someone just watches the video and later types in your shop’s address, it counts as a direct visit there. According to Google, direct visits get no credit in any model unless the entire path consists of them. Discount codes per creator and a “How did you hear about us?” question after checkout pick up part of this.
The same campaign by creator: who brings reach and who sells
| Creator (assumed) | Cost | Reach / orders | CPO |
|---|---|---|---|
| A | €600 | 50,000 / 8 | €75 |
| B | €350 | 20,000 / 20 | €17.50 |
| C | €250 | 10,000 / 12 | €20.83 |
This is the campaign from the worked example, split across the three creators (cost = fee plus products). A brings the most reach and the most expensive orders, B the most orders with less than half the reach. The €30 average CPO hides this. You only get this split if each creator has their own link or code.
A table like this needs creators who actually post. If creators can pick your products in an affiliate programme, that alone does not produce a single post, so there are no figures either. Many partners never become active, and creators that brands would most like to work with rarely post without a reason. The real task is getting many creators to post at the same time and then seeing, creator by creator, who sells.
How it works on wyrld
With wyrld Boost, metrics like the ones in the table above become the basis for the payout: before the start you set which KPIs the pool is paid out on, for example sales, clicks, watch time or reach in Black Friday week. You provide a campaign budget and choose the products and period. Many creators can take part at once, and each one decides whether to join. wyrld measures clicks through the links. Sales count where the shop reports them. Creators upload watch time and reach from their insights. After the campaign you see which creator delivered which results. You decide where people buy, for example only in your own shop. Creators label campaign videos as advertising.
Without a campaign, the brand and shop subscription shows you which creators feature your products, who brings clicks and who sells. For creators who link through wyrld, your affiliate programme usually does not give you a row per creator, because wyrld shows up there as a single partner. With the creators who sell, you can set up your own terms, such as a higher commission or a bonus.
Your products are connected through your existing affiliate programme with one of the major networks, where creators can usually pick them already, or directly with wyrld. Pricing for the subscription and Boost is available on request. We are glad to talk through which should come first for you: Boost, the subscription or connecting your products.
Frequently asked questions.
What is a good engagement rate for influencers?
There is no figure that holds everywhere. The rate depends on what it is divided by: followers, reach or views. It also varies by platform, format and account size. So compare creators on the same platform using the same formula, ideally against their own earlier posts as well.
What is the difference between ROAS and ROI?
ROAS divides revenue by the cost of the campaign. ROI works with profit: what is left after the cost of goods and the campaign cost, divided by that cost. A ROAS of 2 can therefore still mean a loss if your margin is below 50%.
How long after an influencer campaign should I evaluate it?
Do a first report after one to two weeks and the final one once withdrawals and returns are over. In Germany, consumers who order online can usually withdraw within 14 days of delivery (§ 312g(1), § 355(2), § 356(2) BGB), then have up to 14 days after withdrawal to send goods back (§ 357(1) BGB). That is a good four weeks after the last delivery, more with longer returns.
What figures should I get from creators after a campaign?
Reach, views and interactions sit in the creators’ insights, not in yours. So agree in the briefing which figures you will receive and after how many days, for example as a screenshot or an export. Clicks and orders you measure yourself through links, codes and your shop.
Is follower count a good KPI?
Followers show how big an account is, not how many people see a particular post or buy afterwards. For a campaign, the reach of the individual post, the clicks and the orders tell you more. Follower count is most useful when you first shortlist creators.
Last updated:
Sources
The figures in the worked example and in the split by creator are assumptions, not industry benchmarks. The details on Google Analytics and Google Ads come from their help pages, read on 22 October 2026, and may change. This page explains metrics, withdrawal periods and the artists’ social security levy in general terms only and is not legal advice. Accessed on :
- Google Analytics Help: Get started with attribution
- Google Analytics Help: Select attribution settings
- Google Analytics Help: Engagement rate and bounce rate
- Google Analytics Help: Mark events as key events
- Google Ads Help: Clickthrough rate (CTR)
- Google Ads Help: About Target ROAS bidding
- Künstlersozialkasse (artists’ social insurance fund): FAQ for companies, question 24 on influencers (in German)
- German Civil Code (BGB), § 312g (right of withdrawal for distance contracts)
- German Civil Code (BGB), § 355 (right of withdrawal)
- German Civil Code (BGB), § 356 (start of the withdrawal period)
- German Civil Code (BGB), § 357 (legal consequences of withdrawal)