Influencer Marketing ROI: How DTC Brands Measure It

Influencer marketing return on investment (ROI) is the contribution profit a creator deal returns, after product, shipping, fees and returns, divided by what the deal cost: (contribution − cost) ÷ cost × 100. The common form (revenue − cost) ÷ cost × 100, which Sprout Social's guide gives, overstates return because revenue ignores order costs.

For DTC brands, we replace revenue with contribution: what an order leaves after product cost, shipping, payment fees and returns. Every figure from here on comes from one illustrative example, so each number can be traced.

In this illustrative example, revenue is $7,760 (97 orders × $80) against $4,450 of total cost ($2,950 creator cost plus $1,500 media). Revenue-based ROI is (7,760 − 4,450) ÷ 4,450 = 74%, a 1.74x multiple. At a 52% margin you need 1.92x (1 ÷ 0.52) just to break even. Counted on contribution, the 97 orders return $4,035.20 (97 × $41.60) against $4,450, a net of -$414.80, or -9.3%.

Why a DTC creator post earns twice

A creator post earns once as organic content, seen by the creator's own audience. It earns again when you run it as a partnership ad, which runs from the creator's handle and needs the creator's permission (Meta Business Help Center). One fee funds both returns, but they behave differently. Organic orders arrive through links and codes, while paid orders come with their own media cost. One blended number hides which half paid back, and counting the fee in both halves counts the deal twice. For the strategy behind combining creator and paid, read our creator and paid ad hybrid model post.

Count everything you spent to get the post and the right to use it. Product you send creators is counted at your cost of goods sold (COGS), not retail price, so costs match the contribution approach. In the illustrative example, the total is $2,950:

• Fees: $2,000 for the creator's work.
• Product: $100, at your cost, for the items sent.
• Shipping: $50 to send them.
• Usage rights: $500 for the right to run the post as an ad.
• Management time: $300, six hours at an assumed $50 an hour.

Paid media is not on this list. It belongs on the paid side, $1,500 here, so it is never counted twice. A discount given through a creator's code is also a cost, which the example leaves out to keep the arithmetic readable. Management time is the line most brands skip, and skipping it makes small deals look better than they are.

Count orders from first-time customers that arrive through the creator's tracked link, code or survey answer, remove duplicates, and multiply by contribution per order. In the example, 52 orders × $41.60 = $2,163.20 against $2,950 of cost. To track influencer sales without promo codes, tracked links and a post-purchase survey do most of the work.

Links, codes and new-customer rate

Give each creator a link with UTM tags, which are short labels added to a URL that tell your analytics where a visit came from, plus a code. Add a "How did you hear about us?" question after purchase to catch buyers who saw the post and came back another way. Then count only first-time customers. A repeat customer was not acquired by the creator, and counting them inflates the return. Merge the three sources by order number so each order counts once.

Why discount codes undercount

A code only records buyers who typed it. Shoppers who saw the post and bought later, switched devices or skipped the code are missed. In the example, 30 of 52 tracked orders used the code, or 58%. That figure measures the share of tracked orders that used a code, not the creator's total true impact. Code-only reporting would credit the creator with 30 × $41.60 = $1,248 of contribution instead of $2,163.20.

Codes can also overcount: a code posted on a coupon site can collect orders the creator never influenced. Work out your own code share each quarter (code orders ÷ all tracked orders). If it is low, never judge a creator on codes alone.

Score a creator post run as a partnership ad on its cost per acquisition (CPA, media spend divided by purchases) against your breakeven and your normal prospecting ads, which target people who have not bought from you. In the example, $1,500 ÷ 45 purchases = $33.33.

Partnership-ad CPA vs your prospecting CPA

Compare $33.33 with two lines. The first is $41, the CPA of your studio-made prospecting ads. The second is $41.60, the contribution per order, which is the most you can pay for a first order without losing money. The ad clears both. Its 45 orders × $41.60 = $1,872 of contribution against $1,500 of media, a net of +$372.

Before you count those 45 orders, remove any already counted in the organic 52. A single order can carry the creator's link and an ad click. A cheaper CPA than your prospecting ads also does not prove the ad caused the purchases, which the lift check below addresses.

Hook ranking and how to read a creator test are covered in our guide to ranking creator hooks before you scale partnership ads. Our guide to which Facebook ads metric answers which question shows how to read cost per purchase against break-even, hook rate, link click-through rate and ROAS (return on ad spend).

Credit the difference between what the same purchases would have cost at your studio-creative CPA and what they cost with the creator asset. In the illustrative example, 45 purchases at $41 would have cost $1,845. They cost $1,500, so the content credit is $345.

The credit is a comparison with your studio ads on the same 45 purchases, not profit earned, so it is not added to the net. The paid row already books the saving, because its contribution is set against only $1,500 of media. If the creator asset costs more per purchase than your studio creative, the credit is zero and the gap shows in the paid net. Use your own studio CPA, and set it over a period when your studio ads ran at comparable spend. These numbers are illustrative.

For what studio and creator video production costs, read our guide to what a winning ad costs and who makes it.

Pick a window long enough to catch most of your buyers' decisions, then sanity-check the paid result with a simple holdout or branded-search comparison. The window decides which orders count. The lift check asks whether the ads added orders you would not have received anyway.

Match the window to price point

Cheap, impulse purchases convert quickly, while higher-priced or considered purchases take longer. Pull the days between each buyer's first tracked visit and first order from your store data, then choose the shortest window that covers most of them. Apply the same window to every creator so scorecards stay comparable. Our guide to fixing one attribution window per creative test shows how to set it in Meta.

Run a simple lift check

A holdout withholds the ad from part of your audience or a set of regions for a period and compares orders. Meta's own holdout tests work the same way, comparing people who could see the ads with a control group that was intentionally withheld. A branded-search check compares searches for your brand name in the weeks the creator ad runs against the weeks before.

Both are rough. Promotions, seasonality and other channels move the numbers. Treat the result as a sanity check, not proof of causation. If orders barely move when the ad is off, trim the paid credit. If they drop, your paid CPA is more believable.

There is no universal good influencer marketing ROI. Breakeven is creator cost divided by contribution per order, which is 71 orders for the $2,950 of creator cost in this illustrative example. With $1,500 of paid media included, total cost is $4,450 and breakeven is 107 orders.

What the published averages say

Sprout Social reports a Tomoson study finding businesses earn an average of $6.50 for every $1 spent, with the top 13% earning $20 or more. Shopify's guide reports Stack Influence's 2025 benchmarks at $5.20 per $1, while Later gives the same $5.20 average and credits Influencer Marketing Hub. One figure with two named sources is a sign these averages are secondary and unaudited. None reflects your margin, so none tells you whether your creator spend paid back.

How to set your own breakeven

Start with contribution per order: price − product cost − shipping − payment fees − returns allowance. Here, $80 − $24 − $8 − $2.40 − $4.00 = $41.60. Divide creator cost by that figure.

• Creator cost only: $2,950 ÷ $41.60 = 70.9, so 71 orders.
• Total cost, creator cost plus paid media: $4,450 ÷ $41.60 = 107 orders. The example has 97, which is why it ends short.

The same logic gives a revenue multiple: 1 ÷ 0.52 = 1.92x. Our guide to setting a ROAS line from margin covers the margin logic for ad spend. We would treat first-order breakeven as the floor, and count repeat purchases only after you have observed them. A good ROI is a clear margin above that floor.

One table scores a creator across both halves of the deal. The figures reuse the illustrative example's inputs.

Organic alone does not repay the deal, and paid does repay its own media. The whole deal is short of breakeven by $414.80. The creator ad beat the studio-ad CPA by $345 on the same 45 purchases, which is why the creator asset is worth keeping even though the deal is short. In this example, we would renew only if the next deal lowers cost or lifts orders.

Change one input at a time, because the scorecard shows which one is holding the deal back. In the illustrative example the deal is short by $414.80, and each lever closes that gap differently.

• Lift orders. Ten more orders, 107 against 97, cover the shortfall. Ask for a second post or a link in the creator's bio, and score both against the same window.
• Cut cost. Lowering the $4,450 total by $414.80 reaches breakeven on the same 97 orders. Renegotiate the fee or the $500 of usage rights, and pay for usage rights only if you will run the post as an ad.
• Raise contribution per order. On the same 97 orders, breakeven needs about $45.88 per order, $4.28 more than $41.60. Shipping, bundling and returns policy move that figure, not the creator.
• Trim paid credit. If the lift check shows the ad added fewer orders than the paid row counts, cut the paid media and score the organic row alone.

A creator who needs all four changes before the deal pays back is a creator to stop renewing.

Build the scorecard for your last three creators before you renew any of them. Start with your own contribution per order, add every cost from the list above, score organic and paid in separate rows, and add the content credit last. Renew creators who clear breakeven and rework the terms of those that do not.

Per-creator, data-backed measurement matters on Meta now because the same creator asset runs as organic content and as a partnership ad, and budgets are decided creator by creator. A blended number hides which creators earned their fee.

All dollar figures, order counts, breakeven values and percentages in the example are our own illustrative model, built from the inputs table.