UGC ROI and CPM Benchmarks: What to Measure in 2026

The three-layer UGC scorecard

UGC, user-generated content, here means short videos that creators film and post on their own accounts for a brand. Most weak UGC reports either stop at the first layer or jump straight to the third. A report that stops at cost per view rewards bought or accidental views. A report that only looks at sales punishes content that built awareness which paid off weeks later. Report all three layers, in that order, for every creator and every platform, and the arguments about whether "UGC works" get much shorter.

Set a target for each layer before the campaign starts, not after the numbers arrive. For the first layer that is a view target and a ceiling on effective CPM. For the second it is a floor, such as the completion rate of your own best organic videos on the same platform. For the third it is the number of orders or leads that would make the campaign break even, which the formulas section below shows how to work out. Targets written down in advance stop a report from being reverse-engineered to look like a success.

What counts as a view on Reels, TikTok and Shorts?

Before comparing CPMs across platforms, check that the denominator means the same thing. It does not, and the differences are large enough to move a budget decision.

Two consequences follow. First, Instagram and YouTube Shorts count replays, so a short video that loops can record more views than it has viewers. Where a platform reports reach, the ratio of views to reach tells you how much of the total is repeat viewing. Second, YouTube's own counts are not comparable across time: Shorts views before and after 31 March 2025 use different rules, and the August 2026 change extends the start-to-play rule to every format. Any year-on-year YouTube comparison should use engaged views.

A practical rule for reports: record which definition each number uses, where it came from and the date it was read. A screenshot of the platform's insights panel with the date and metric name visible settles most disputes about a number's origin. Use the creator's own analytics for organic posts and Ads Manager definitions for anything you paid to promote, and never add the two together in one column.

TikTok's advertising metrics are the strictest by default. A play counts once per impression, replays are excluded, and a 6-second focused view needs at least 6 seconds of watching, a full play of a shorter video, or an interaction within the first 6 seconds^[3]. If one platform looks much cheaper than another, check the definition before moving money. Our short-video platform guide compares audience reach by market.

What is a good CPM for UGC in 2026?

CPM, cost per mille, is the cost of 1,000 views. For creator content the useful version is effective CPM: everything you paid, divided by verified views, times 1,000. It lets a flat-fee creator deal, a view-priced marketplace and a paid social campaign sit on the same line of a spreadsheet.

Two features of this ladder matter for ROI. The rate is flat: once total views land in a band, every view is billed at that band's price, not only the views above a threshold. And the deposit, set at 115% of the target fee, is the most you can be charged; unused deposit returns to your credit when the campaign closes. Totals below 100,000 views are billed at US$2.90 per 1,000, and there is no top-up fee.

Effective CPM also shows where the risk sits. With a flat creator fee, the brand pays the same whether the video reaches a few thousand people or several hundred thousand, so the effective CPM is only known afterwards. With view-based pricing, cost rises and falls with delivery, and the remaining uncertainty is which band the total will land in. Neither model is automatically cheaper. The point is to compare them on the same unit once the results are in, and to know in advance which risk you are carrying.

So what counts as good? Meta and TikTok do not publish average CPMs for Taiwan, Hong Kong, Malaysia or Singapore, and third-party averages mix objectives, formats and seasons into one number. The honest benchmark is your own. Take the CPM from the last three months of paid social in each market, straight from Ads Manager, and put it next to the effective CPM of your creator programme, measured on a comparable view definition. For flat-fee creator rates, see what a UGC video costs.

Worked example: a 500,000-view target

Read the rows as a range, not a forecast. When content over-delivers, the deposit cap works in the brand's favour: in the third row, 640,000 views at US$2.00 would be US$1,280, but the charge stops at the US$1,150 deposit, so the effective CPM drops to US$1.80. When content under-delivers, as in the first row, the total lands in a higher band and the rate rises to US$2.30, although the bill is still lower. New approved brands get US$50 of trial credit, which is enough to run the whole flow on a small target before committing a full budget.

Which quality signals are worth tracking?

Cheap views only help if people watched. Three quality signals are worth tracking for every video: completion, meaning full plays as a share of views; early retention, meaning 2-second or 6-second views on TikTok and engaged views on YouTube; and deep engagement, meaning shares and saves rather than likes. Likes are cheap to give; a share puts the viewer's own reputation behind the video.

Nielsen's Brand Impact norms for Q1 2022, based on around 200 campaigns, also found that influencer ads lifted brand affinity and purchase intent by 9 points on average^[7]. That is a survey of people who saw the ads, not of the whole market, but it is the right kind of evidence for the second layer: did the video change what viewers think of the brand?

Use quality signals to diagnose, not only to grade. A weak 2-second or 6-second view rate usually points to the opening: the hook, the cover frame or the first line. Good early retention with poor completion suggests the middle drags or the product appears too late. Strong completion with few shares or saves suggests the video was watchable but gave viewers no reason to pass it on. Compare creators within a platform, never across platforms, because the definitions differ.

Platform studies point the same way, with a caveat attached. TikTok's Spark Ads playbook reports that Spark Ads, which run from a creator's organic post, had a 30% higher completion rate, a 142% higher engagement rate and a 43% higher conversion rate than non-Spark in-feed ads, using TikTok internal data from Q4 2020 to Q2 2021^[8]. Treat vendor numbers like these as direction, not as a forecast for your brand: they are platform-reported and several years old.

How do you connect UGC to sales?

Organic creator videos rarely carry a clickable link in the video itself, so attribution has to be planned before the first post goes live, not reconstructed afterwards.

Expect under-counting. Some viewers search for the brand instead of tapping a link, some screenshot the product, and some buy weeks later. A useful cross-check is to compare branded search and direct traffic in each market for the two weeks after a wave of posts with the two weeks before. The difference is not proof, but a consistent pattern across several waves is hard to ignore.

The strongest evidence is a holdout. Run creators in one market or city and not in a comparable one for the same weeks, then compare sales growth between the two. It is not a laboratory experiment, because markets differ, but it answers the question finance actually cares about: would these sales have happened anyway? In a multi-market campaign, also put the market in utm_campaign and convert revenue into one reporting currency on a stated date, so orders in ringgit and New Taiwan dollars can sit in one total.

Measurement gets cleaner when creator videos are reused as ads. Meta partnership ads and TikTok Spark Ads report clicks and conversions in Ads Manager like any other ad. Keep organic and paid results in separate columns, because TikTok attributes all Spark Ads engagement, including views, to the original organic post^[9]. Otherwise the same view can be paid for twice.

ROI and ROAS formulas, with a break-even check

ROAS, return on ad spend, is attributed revenue divided by cost. ROI, return on investment, is profit divided by cost. They answer different questions. ROAS says how much revenue each dollar brought in; ROI says whether the business is better off after paying for the goods and the campaign.

Use gross profit, not revenue, when you calculate ROI; a healthy-looking ROAS on a low-margin product can still lose money. Include every cost, too: creator or platform fees, product samples and shipping, agency or staff time, and any paid amplification. The break-even row is the fastest sanity check before launch. If the number of orders you would need looks implausible for the audience size, change the offer or the target before you spend.

UGC also leaves assets behind. If a creator video is licensed for ads, some teams add the production cost they avoided as a "content value" line. Keep it separate from ROI: it is only a real saving if you would otherwise have paid for that production, and only once the usage rights are secured. Our guide to disclosure and usage rights explains the difference between posting rights and ad rights.

Why only verified views belong in ROI

Every ROI figure above is only as good as the view count underneath it. Bought views, bot traffic and views counted too early all inflate the first layer and flatter the whole report. Singapore's advertising guidelines explicitly prohibit boosting engagement through fraudulent means, naming bulk-bought likes, fake accounts and programmes that generate page views^[11].

Clipy builds three controls into billing. Views count only after a video has been live for 7 days, counted from its post date, and creators confirm that their numbers are final when they submit. Admins then review the data. For Instagram, views are adjusted for natural traffic, with feed and profile traffic capped, which removes the incentive to buy views. The result is a smaller but more defensible number, which is the number a finance team will accept.

What should a brand ask for? A screenshot of each video's insights taken after day 7, showing the date, the view count and, on Instagram, where the views came from. Ask every creator for the same evidence so the numbers are comparable, and store it next to the post URL. On Clipy, creators submit their numbers and confirm they are final, and the numbers are verified — automated checks, then our team for anything held back — before they count towards billing.

A one-page monthly UGC report

Sort the sheet by effective CPM, then read across. The creators worth rebooking are usually in the top half on cost and the top half on completion or shares, even when their follower counts are modest. Creators who are cheap per view but weak on completion need a sharper hook in the first seconds, not a bigger budget.