💬 Quick answer: For DTC ecommerce, the best-disclosed TikTok panel — 5,900+ brands over twelve months — reports a median CPM of $4.08, CTR 0.61%, conversion rate 1.56%, ROAS 1.51, cost per acquisition $17.07 and average order value $50.32. TikTok impressions cost roughly a third of Meta’s and a quarter of Google’s, but click through at about a quarter of Meta’s rate. These are third-party figures, not ours — our own TikTok panel is too small to publish medians from, and we say so below rather than dressing it up.
Updated 26 September 2026 · Data window: September 2025 – August 2026 · Source: vetted third-party panels, plus a disclosed first-party cohort
TL;DR
- We do not have a publishable first-party TikTok panel. Only one of our TikTok ad accounts clears the $10,000 spend floor we use across this series, and it holds 94% of the cohort’s spend. A one-account median is not a benchmark, so we do not publish one.
- Instead this page reports the two TikTok datasets that disclose a panel size, a window and a segment, and audits them: Triple Whale (5,900+ brands) and Two Minute Reports (2,500+ advertisers).
- Headline third-party figures: CPM $4.08, CTR 0.61%, CVR 1.56%, ROAS 1.51, CPA $17.07, AOV $50.32.
- The ROAS row does not survive a consistency check. ROAS × CPA should equal AOV; on TikTok that gives $25.78 against a stated $50.32 — a 2× gap that the same provider’s Google figures do not have. The most likely cause is incomplete purchase tracking on TikTok.
- TikTok is not simply “worse” than search. It is a different economic shape: the lowest CPM, the lowest cost per purchase and the lowest order value of the three platforms in this series.
- Our own cohort, published as an observation rather than a benchmark: 6 accounts, $66,974 spend, pooled CPM $4.66, CPC $0.55, CTR 0.84%, implied order value $50.69.
What is a good TikTok ads CPM for ecommerce?
The best-disclosed panel — 5,900+ ecommerce brands over twelve months — reports a median CPM of $4.08, with an industry range of $3.46 to $4.61. That is roughly a third of the $13.52 median our own first-party Meta panel pays and a quarter of the $16.76 our Google panel pays.
What is a good ROAS for TikTok ads?
Published panels put TikTok ROAS at 1.51, against 2.96 for Meta and 4.28 for Google in our first-party data. Most of that gap is order value rather than media performance: TikTok-attributed orders average $50.32 against $121 to $145 on search and feed. Treat 1.51 as a floor to interrogate, not a target, because TikTok purchase tracking is measurably less complete than the other platforms’.
Is TikTok advertising cheaper than Meta?
Per impression, clearly yes: about $4.08 CPM against $13.52. Per click the advantage mostly disappears, because TikTok click-through rates run around 0.61% against Meta’s 2.68%, which puts TikTok CPC at roughly $0.67 against Meta’s $0.57. Cheap impressions do not automatically mean cheap customers.
Why this page is not first-party
Every other benchmark in this series is built from our own client data. This one is not, and the reason matters more than the numbers.
We queried every TikTok ad account we manage for the twelve months to August 2026. Six returned spend. Total spend across all six was $66,974, and one account accounted for 93.9% of it. The other five spent between $250 and $1,319 each — enough to learn from, nowhere near enough to benchmark with.
The panel rule we apply across this series is a $10,000 minimum spend in the window, so that a single account’s early-learning-phase numbers cannot masquerade as a benchmark. Exactly one TikTok account clears it. A median of one is not a median, and publishing one dressed as a benchmark is how this genre goes wrong.
So this page does something more useful than inventing a panel: it takes the two TikTok datasets that actually disclose their methodology, checks their internal consistency, names the ones we rejected, and adds our own cohort as a clearly-labelled observation. When our TikTok spend crosses the floor on enough accounts, this page becomes first-party like the rest.
The headline numbers
From the largest disclosed TikTok ecommerce panel — 5,900+ brands, August 2025 to July 2026:
| Metric | Median | Industry range | Source |
|---|---|---|---|
| CPM | $4.08 | $3.46 – $4.61 | Triple Whale, 5,900+ brands |
| CTR | 0.61% | 0.41% – 0.74% | Triple Whale, 5,900+ brands |
| Conversion rate | 1.56% | 0.38% – 2.08% | Triple Whale, 5,900+ brands |
| ROAS | 1.51 | 0.02 – 2.70 | Triple Whale, 5,900+ brands |
| Cost per acquisition | $17.07 | $5.46 – $28.90 | Triple Whale, 5,900+ brands |
| Average order value | $50.32 | $0.38 – $73.79 | Triple Whale, 5,900+ brands |
| CPC (derived) | $0.67 | — | CPM ÷ 1,000 ÷ CTR |
And from the second disclosed panel — 2,500+ advertisers, retail and ecommerce, January to December 2025: median CPC $0.26, CTR 0.82%, CPA $3.78, conversion rate 2.83%, on a median advertiser spend of $809.
Two things to notice before using any of this. First, the two panels disagree on CPC by roughly 2.5× ($0.67 derived against $0.26 reported), which is what you would expect when one panel is ecommerce brands and the other is advertisers of every size with a median spend under a thousand dollars. Second, one row in the industry ranges is not a real number: an average order value of $0.38 and a ROAS of 0.02 in a single category is a tracking artefact, not a business. Treat by-industry cuts from any provider with that in mind.
The two panels we use, and the ones we rejected
A benchmark page is only as citable as its weakest source. We admit a source only if it publishes all three of the following. Anything missing one is excluded, however confident it looks:
- A panel size — the number of brands, advertisers or accounts behind the figures.
- A time window — the period the data covers, not just a publication date.
- A segment — the industry, geography and campaign type it applies to.
| Source | Panel | Window | Verdict |
|---|---|---|---|
| Triple Whale | 5,900+ ecommerce brands | Aug 2025 – Jul 2026 | Admitted |
| Two Minute Reports | 2,500+ advertisers on its reporting platform, 28 industries | Jan – Dec 2025 | Admitted, segment breadth disclosed |
| Lebesgue | none disclosed | none disclosed | Rejected — describes its data only as “ecommerce advertising data analysed by Lebesgue”, with no advertiser count and no date range, in a report labelled as current |
| Nine further “TikTok benchmark” pages ranking for these terms | none disclosed | none disclosed | Rejected — figures recirculated from other publishers with no original panel. Where a page cites another publisher’s data, we cite that publisher directly instead |
That last row is the honest state of this topic. Most results for “TikTok ads benchmarks” are the same handful of numbers passed between content pages, each one acquiring a little more confidence and a little less provenance on the way. Before using any figure, open the methodology section. If it cites other publishers rather than its own data, go to them.
Auditing the ROAS figure
There is a simple identity that any benchmark publishing ROAS, cost per acquisition and average order value must satisfy:
- ROAS × cost per acquisition = average order value, because revenue ÷ spend × spend ÷ purchases leaves revenue ÷ purchases.
Applied to the same provider’s two platform reports, over the same window and with the same methodology, it behaves very differently:
| Panel | ROAS | CPA | Implied AOV | Stated AOV | Consistent? |
|---|---|---|---|---|---|
| Google Ads, 21,000+ brands | 3.27 | $28.14 | $92.02 | $87.65 | Yes — within 5% |
| TikTok, 5,900+ brands | 1.51 | $17.07 | $25.78 | $50.32 | No — 2× apart |
Some slack is expected, because each metric is an independent median across brands and medians do not compose. But a 5% gap on Google and a 95% gap on TikTok, from one provider using one method, points at something specific to TikTok rather than to the arithmetic.
The most likely explanation is incomplete purchase tracking. TikTok attribution is widely understood to under-capture conversions relative to Google and Meta — weaker pixel coverage, shorter click windows, more in-app and Shop journeys that never reach the site tag. If a meaningful share of a panel’s accounts record only part of their TikTok revenue, ROAS is pushed down while a separately-computed average order value, drawn from the orders that are captured, stays roughly right.
The same pattern appears in the second panel: a median CPC of $0.26 and a conversion rate of 2.83% imply a cost per acquisition of $9.19, against the $3.78 reported. Again, medians do not compose — but again the direction is the same.
Our reading, stated as a judgement rather than a finding: published TikTok ROAS figures are likely to be a floor rather than a central estimate, and TikTok’s real contribution is probably better than 1.51 suggests. That is not a reason to ignore the number; it is a reason to measure TikTok with blended MER or incrementality rather than platform ROAS. The MER of 0.23 that the same panel publishes — TikTok spend as a share of total revenue — is less exposed to this problem.
What our own accounts show
Published as a cohort observation with every limitation on the table. This is not a benchmark and should not be cited as one.
| Metric | Our cohort (pooled) | Triple Whale (5,900+ brands) | Difference |
|---|---|---|---|
| CPM | $4.66 | $4.08 | 14% higher |
| CPC | $0.55 | $0.67 (derived) | 17% lower |
| CTR | 0.84% | 0.61% | 38% higher |
| Cost per purchase | $5.52 | $17.07 | 68% lower |
| Purchase ROAS | 9.19 | 1.51 | 6× higher |
| Implied average order value | $50.69 | $50.32 | 0.7% higher |
Only the last row means anything. Our implied order value of $50.69 lands within a percent of the 5,900-brand panel’s $50.32 — two entirely independent datasets agreeing on what a TikTok-attributed order is worth. That is a genuine corroboration, and it is also modest supporting evidence for the audit above: the stated average order value looks reliable, while the ROAS row looks understated.
Every other row is dominated by one account:
- 6 accounts, $66,974 total spend, 14,367,987 impressions, 120,712 clicks, 12,142 purchases.
- One account holds 93.9% of the spend and almost all of the purchases. The pooled ROAS of 9.19 is essentially that one account’s result.
- The other five spent $250 to $1,319 each — below any sensible learning-phase threshold, so their rates measure ramp-up rather than steady state.
- Because only one account clears our $10,000 floor, no median, interquartile range or percentile is published. Pooled figures only, and pooled figures describe the biggest account.
If you want to see how differently that reads when a panel is properly sized, our Meta Ads benchmark covers 15 accounts and our Google Ads benchmark covers 28, both with full distributions.
TikTok versus Meta and Google
Setting the disclosed TikTok panel beside our own first-party Meta and Google panels gives the cross-platform picture. The TikTok column is third-party; the other two are ours. They are not the same brands and the windows differ by up to two months, so read the shape rather than the decimals.
| Metric | TikTok (5,900+ brands, third-party) | Meta (15 brands, ours) | Google (28 brands, ours) |
|---|---|---|---|
| Median CPM | $4.08 | $13.52 | $16.76 |
| Median CPC | $0.67 (derived) | $0.57 | $1.11 |
| Median CTR | 0.61% | 2.68% | 1.79% |
| Median ROAS | 1.51 | 2.96 | 4.28 |
| Median cost per purchase | $17.07 | $49.04 | $32.47 |
| Average order value | $50.32 | $145.16 | $121.40 |
Three things fall out of that table, and the first is the one most often missed:
- Cheap impressions do not survive into cheap clicks. TikTok CPM is 70% below Meta’s, but TikTok CTR is 77% below Meta’s too. The two nearly cancel, leaving TikTok CPC slightly above Meta’s. Anyone justifying a TikTok budget on CPM alone is arguing from the wrong metric.
- The ROAS ranking is largely an order-value ranking. TikTok $50, Google $121, Meta $145 — the same order as the ROAS column once you account for tracking. TikTok sells smaller baskets, and a smaller basket produces a smaller multiple on the same media efficiency.
- TikTok has the lowest cost per purchase of the three. $17.07 against $32.47 on Google and $49.04 on Meta. That is the row TikTok wins, and it is consistent with cheap impressions plus small baskets: cheap customers buying cheap things.
The shape of TikTok economics
Put together, the disclosed data describes a channel with a distinctive and internally coherent shape rather than a weak version of Meta:
| Characteristic | TikTok | Consequence |
|---|---|---|
| Impression cost | Lowest of the three, ~$4 CPM | Reach is cheap; awareness maths works easily |
| Click rate | Lowest of the three, ~0.6% | Most of that cheap reach never clicks |
| Order value | Lowest of the three, ~$50 | Every downstream revenue metric is scaled down |
| Cost per purchase | Lowest of the three, ~$17 | Acquisition is genuinely cheap in absolute terms |
| Platform ROAS | Lowest of the three, ~1.5 | Partly real, partly a tracking artefact |
| Measurement quality | Weakest of the three | Judge with blended MER, not platform ROAS |
The practical implication: TikTok is a volume-and-discovery channel whose contribution shows up off-platform. A brand with a $50 basket and a $17 acquisition cost has a working channel; the same brand reading a 1.5 platform ROAS against a 4.0 search ROAS will conclude it does not, and turn off the channel that was feeding search.
That is the same trap our returning vs new customer benchmark documents from the other direction: returning visitors convert 2.60× better than new ones across every store we measured, so any channel credited with returning traffic looks strong and any channel that creates first-time demand looks weak. Discovery channels are systematically under-credited by last-touch measurement.
What to do with this
- Do not justify TikTok on CPM. Compare CPC and cost per purchase, where the advantage is real but much smaller than the CPM gap implies.
- Judge TikTok on blended MER, not platform ROAS. Total revenue divided by total ad spend, before and after a TikTok budget change, is the only measurement not exposed to TikTok’s attribution gap. The disclosed panel puts TikTok MER at 0.23.
- Check your own order value first. If your basket is well above $50, the published TikTok ROAS and CPA figures do not transfer to you at all — multiply the ROAS by the CPA of any benchmark to see what order value it assumes.
- Run a holdout or geo test before scaling or cutting. On a channel whose platform numbers are demonstrably inconsistent, incrementality is worth more than another week of dashboard staring.
- Verify your pixel and events coverage before believing a low ROAS. Much of the gap identified above is measurement, and some of it will be yours.
- Treat any TikTok benchmark without a disclosed panel as unusable, including the ones that agree with you.
Methodology and limits
This page reports other people’s panels and says so in every row. Figures are reproduced as each source publishes them, with the source, panel size and window stated alongside. Where we derive a figure — TikTok CPC from CPM and CTR, implied average order value from ROAS and cost per acquisition — the derivation is shown so it can be checked. We never average a first-party figure together with a third-party one.
Our first-party cohort: TikTok Ads data for every client ad account we manage, for the trailing twelve months from 1 September 2025 to 31 August 2026. Six accounts returned spend; all six are included in the pooled figures, because excluding the small ones would leave a single account. Pooled means total spend divided by total impressions, clicks or purchases — not a median. Attributed revenue is derived per account as ROAS × spend and summed.
Known limits, stated plainly:
- This is not a first-party benchmark. The headline figures belong to Triple Whale and Two Minute Reports, not to us. We have audited them, not reproduced them.
- Our cohort is n = 6 with 94% concentration. No median, no interquartile range, no percentile is published from it, and its ROAS should be read as one account’s result.
- Windows differ between sources by up to twelve months, and TikTok auction prices move quickly. Each row states its own window.
- The cross-platform table mixes panels. The TikTok column is 5,900+ third-party brands; the Meta and Google columns are 15 and 28 of our own accounts. Different brands, different sizes, windows up to two months apart.
- TikTok attribution is the weakest of the three platforms, and the audit section quantifies how much that matters. Every revenue-derived TikTok figure on this page inherits that weakness.
- Segment breadth differs. Triple Whale’s panel is ecommerce brands; Two Minute Reports’ covers 28 industries with a median advertiser spend of $809, which makes its absolute cost figures hard to read across to a DTC brand at scale.
- TikTok Shop and in-app purchase journeys are not separated out in any of these datasets, and they behave differently from site conversions.
How to benchmark your own account
- In TikTok Ads Manager, set the date range to the last full 12 months and read spend, impressions, clicks, complete payments and complete payment ROAS at account level.
- Place CPM first against ~$4.08, then CTR against ~0.61%. Those two are the least exposed to attribution problems and tell you whether the media and creative are working.
- Then cost per purchase against ~$17.07 — but only if your order value is near $50. If it is not, this row does not apply to you.
- Compute your own implied order value as ROAS × cost per purchase and compare it to your true AOV from your ecommerce platform. A large gap is a tracking problem, and it is worth more attention than the ROAS number itself.
- Ignore platform ROAS as a decision input until you have checked pixel and events coverage.
- Model the budget against a blended MER target rather than a channel ROAS target — our MER calculator and TikTok ads budget calculator both work that way.
Sources and methodology
Frequently asked questions
What is a good TikTok ads CPM for ecommerce?
The best-disclosed panel — 5,900+ ecommerce brands, August 2025 to July 2026 — reports a median CPM of $4.08, with an industry range of $3.46 to $4.61. Our own six-account cohort pools to $4.66 over an overlapping window. TikTok impressions cost roughly a third of the $13.52 our first-party Meta panel pays and a quarter of the $16.76 our Google panel pays.
What is a good CPC for TikTok ads?
Derived from the 5,900-brand panel's CPM and CTR, TikTok CPC works out at about $0.67. A second panel of 2,500+ advertisers reports $0.26 for retail and ecommerce, but on a median advertiser spend of only $809, which makes it a different kind of measurement. Our own cohort pools to $0.55. The useful point is that TikTok CPC lands close to Meta's $0.57 despite a CPM roughly 70% lower, because TikTok click-through rates are about a quarter of Meta's.
What is a good ROAS for TikTok ads?
Published panels report 1.51, against 2.96 for Meta and 4.28 for Google in our first-party data. Two adjustments matter before using it. First, order value: TikTok-attributed orders average $50.32 against $121 to $145 on search and feed, and a smaller basket produces a smaller multiple at the same media efficiency. Second, tracking: ROAS times cost per acquisition should equal average order value, and on TikTok that identity is 2 times out, which suggests published TikTok ROAS is a floor rather than a central estimate.
Is TikTok advertising cheaper than Meta or Google?
Per impression, clearly. About $4.08 CPM against $13.52 on Meta and $16.76 on Google in our first-party panels. Per click the advantage nearly disappears, because TikTok's 0.61% click-through rate is about a quarter of Meta's 2.68%. Per purchase TikTok is cheapest again at roughly $17.07 against $49.04 on Meta and $32.47 on Google — but that partly reflects a $50 basket rather than $121 to $145.
Why don't you publish your own TikTok benchmark like your Meta and Google ones?
Because the data does not support one. We queried every TikTok account we manage for the twelve months to August 2026. Six returned spend, totalling $66,974, and one account held 93.9% of it. The panel rule we apply across this series requires at least $10,000 of spend per account in the window, and exactly one TikTok account clears it. A median of one account is not a median, so we publish the cohort as a pooled observation and lean on disclosed third-party panels for the headline figures instead.
Which TikTok benchmark sources can be trusted?
We use a source only if it publishes a panel size, a time window and a segment. Two TikTok datasets do: Triple Whale, with 5,900+ ecommerce brands over August 2025 to July 2026, and Two Minute Reports, with 2,500+ advertisers across 28 industries over calendar 2025. Ten others were rejected, including one that describes its data only as ecommerce advertising data it analysed, with no advertiser count and no date range. Most results for this search term recirculate the same figures without an original panel; where a page cites another publisher, go to that publisher directly.
How should I measure TikTok if its ROAS is unreliable?
Use blended MER — total revenue divided by total ad spend across all channels — measured before and after a TikTok budget change, and back it with a geo or holdout test. Blended measurement is the only approach not exposed to TikTok's attribution gap. The disclosed panel puts TikTok MER at 0.23, meaning TikTok spend equal to about 23% of total revenue. Verify your own pixel and events coverage before concluding the channel does not work.
Why is TikTok's average order value so much lower than Meta's or Google's?
Because of what people buy there. TikTok-attributed orders average $50.32 across 5,900+ brands, and our own cohort independently implies $50.69 — two unrelated datasets landing within a percent of each other. Our first-party Meta panel implies $145.16 and our Google panel $121.40. Lower-priced impulse and discovery purchases dominate TikTok, and that single fact explains most of the gap in every downstream revenue metric.
What was measured, and over what period?
Headline figures come from Triple Whale's panel of more than 5,900 ecommerce brands for August 2025 to July 2026, and from Two Minute Reports' panel of more than 2,500 advertisers for January to December 2025, reproduced as each publishes them. Our own cohort is TikTok Ads data for six client accounts for the trailing twelve months from 1 September 2025 to 31 August 2026 — $66,974 of spend, 14,367,987 impressions, 120,712 clicks and 12,142 purchases, reported pooled rather than as percentiles because one account holds 93.9% of the spend.
Do these figures represent the whole industry?
The third-party panels are large enough to describe the ecommerce advertiser base reasonably well, within the limits of their own disclosure. Our cohort does not represent anything beyond itself, and we publish it as an observation rather than a benchmark. Nothing on this page should be read as a first-party Top Growth Marketing benchmark in the way our Meta, Google, email and conversion rate pages should.
- Triple Whale — TikTok Ads Benchmarks by Industry (5,900+ brands, Aug 2025 – Jul 2026)
- Two Minute Reports — TikTok Ads Benchmarks by Industry (2,500+ advertisers, Jan – Dec 2025)
- Triple Whale — Google Ads Benchmarks by Industry (21,000+ brands), used for the consistency check
- TikTok Business API documentation
- Meta Ads Benchmarks
- Google Ads Benchmarks
- Average Order Value Benchmarks
- Returning vs New Customer Benchmarks
- TikTok Ads Budget Calculator
- MER Calculator
See also our average order value benchmark for why TikTok’s $50 basket drives most of the ROAS gap, and our Meta Ads benchmark for a properly-sized first-party paid social panel.