If you sell on TikTok Shop US where paid placement has got more expensive and the ROAS on your product ads no longer covers the spend, the change that hurt this year was not in your product or in your pricing. It was in where your traffic comes from. The number sellers keep quoting is $1.50, and it is real but it is being read slightly wrong. Current 2026 benchmarks put TikTok CPC somewhere between $0.17 and $1.50 depending on category, with the cross-industry average sitting near $1.02 — health and supplements run $0.60 to $1.50, electronics $0.70 to $1.80, home and kitchen $0.35 to $0.90, beauty and personal care $0.30 to $0.85. So $1.50 is the top of the band, not the middle, and it is the top of the band in exactly the categories where competition for the same audience is fiercest. The more useful reading is seasonal: between October and early December sellers commonly see CPC rise 30 to 60 per cent as holiday competition arrives. That is what is hitting budgets right now, and the response most sellers reach for — bid higher and accept the cost — treats the symptom as the cause.
Run the arithmetic on your own monthly budget before you change anything, because it is usually worse than the feeling. At $0.55 CPC, $5,000 buys roughly 9,090 clicks. At $1.02 — the current cross-industry average — the same $5,000 buys about 4,900. Hold conversion at 2.5% and average order value at $38 and you go from roughly 227 orders and $8,600 of GMV to about 123 orders and $4,700. Nothing in that calculation changed except the price of the click, and it nearly doubled: your product, your listing and your audience are identical on both lines. Now add the seasonal layer on top. A 30 to 60 per cent rise between October and early December puts that $1.02 on a path to somewhere between $1.33 and $1.63 before the year closes, which removes another quarter to a third of the remaining volume. Two responses are obvious — spend more, or accept less — and both are worse than the third, which is paying less per click for the same reach by changing what you are amplifying.
What actually rose: the price of placement, not the price of reach
The distinction that unlocks this is simple and easy to miss. What got more expensive is buying placement for a product listing. What did not get more expensive is reaching people with a video they actually want to watch, and those are two separate purchases. Spark Ads promote an existing organic post — your own, or a creator's with authorisation — carrying the original caption, the creator's profile image, and the likes, comments and shares that post already earned into the paid placement. Because the auction reads existing engagement as a quality signal, Spark Ads typically run 30 to 50 per cent below the CPC of equivalent brand-made in-feed ads, in the $0.20 to $0.50 band, and they have been reported to outperform standard in-feed placements on key metrics by 20 to 40 per cent, with CPM roughly 20 per cent lower. You are not buying cheaper ads. You are buying the same reach at a lower price by attaching it to content that already earned attention for free.
One caveat saves money and it trips up experienced advertisers constantly. If you run GMV Max, which allocates budget automatically across Spark, video shopping ads, product shopping ads and live ads toward a target ROI, do not judge it on CPC. GMV Max intentionally pays more per click for higher-intent users, so a rising CPC under GMV Max is frequently evidence that the algorithm is working rather than failing. The number to hold it to is ROAS and, more honestly, contribution margin after commission, shipping and refunds. Platform-wide ROAS benchmarks sit around 3.5x, with strong beauty and home sellers reporting 4x to 6x; the low end of the range reports CPA of $5 to $15, which is profitable or not entirely depending on your margin structure. Pick the metric that actually decides your case before you start optimising toward one.
Step 1 — Change the job the ad is doing
Stop treating the ad as a way to buy a position for your product page and start treating it as a way to add paid distribution to a video that has already earned organic distribution. In practice this means the sequence flips: the creator publishes first, the video runs free, and the paid spend follows the evidence rather than preceding it. Concretely, set aside the budget you would have put behind a product shopping campaign and allocate it to amplifying creator posts instead. The objection is immediate — you lose the ability to guarantee a position, and you cannot control exactly who sees it. That is real, and it is also the thing you were paying the higher CPC for. If control of placement is worth more to you than cost per outcome, this approach is not for you. For most sellers whose ROAS stopped working, it is not worth more.
Write this down as a rule your team can enforce without you in the room: no paid spend on a piece of creative until it has run organically and been read. It sounds slow and it is the opposite — the screening period costs you two days, while skipping it costs you the full budget on content you have no evidence for. Sellers who adopt this usually find their effective cost per outcome falls even though their cost per click barely moves, because the money that used to go to unproven creative now goes to proven creative.
The screening sheet in step two is the part people skip and the part that does the work. register with DAMI and ask the team for the creative screening sheet and get the creative screening sheet we use — the three organic signals to read at 48 hours, the threshold for each, and a small-budget test plan for three to five pieces.
Step 2 — Screen creative in the free window before you pay for any of it
Give every creator video 24 to 48 hours of organic life before it gets a dollar, then read three signals. Watch-through: is completion holding up relative to the creator's other recent posts, or is this one being abandoned early? Comment volume and type: are people asking where to buy it, tagging someone, or arguing, or is the section quiet? And product-card click rate, which is the closest thing you have to intent before money is involved. Amplification cannot fix a creative problem — putting paid distribution behind a video that is not resonating spends your budget on content that was never going to work, at a larger scale. The free window is the cheapest filter you will ever get, and it costs two days.
Set thresholds before you look, or you will rationalise whatever you see. A workable starting point: completion at or above the creator's own recent median, at least a handful of comments that mention buying rather than only praise, and a product-card click rate above whatever your own catalogue averages. Pieces that clear all three go into the paid test. Pieces that clear two get another 48 hours rather than a decision. Pieces that clear one or none are dropped and, more usefully, examined — if five in a row fail the same signal, the brief is wrong and no amount of budget will rescue it.
Step 3 — Spend small on three to five pieces, then scale on one number
Put a small, equal budget behind three to five pieces that cleared the screen — enough to generate a readable number, not enough to matter if all five fail. Run them for four to five days without touching them, then scale on one number: contribution margin per dollar of ad spend, which is ROAS adjusted for commission, shipping and refunds. ROAS alone will mislead you here, because it is computed on gross and your margin is not. Take the one or two pieces that clear your threshold and increase budget in steps rather than all at once, because TikTok's creative fatigue curve is shorter than any other major paid channel and a sudden jump in spend accelerates it.
Plan for fifteen to twenty distinct concepts before meaningful spend, and a sustained weekly cadence of new variants after that. Automated campaign types consume creative faster than manual ones, and teams that cannot sustain weekly output see performance decay regardless of budget or bidding strategy. This is the part that turns a media problem into an operations problem: the constraint is no longer how much you can spend, it is how much original video you can produce and screen per week. Sellers who solve the production side usually find the media side gets much easier, because they always have something new to test.
Step 4 — Use creator content to amortise what creative costs you
Creator content is the cheapest source of original video you have, and the reason is structural rather than a bargain. You pay commission on outcomes instead of a fee up front, so the cost of a piece that fails is close to zero, and every piece that succeeds also produces the organic signals you need for step two. Compare that with producing brand-made creative: you pay for it whether or not it works, and it arrives with no engagement history for the auction to read. Reported figures put creator-economy ads at roughly 96 per cent higher ROAS than comparable spend on other digital channels, and the mechanism is exactly this — you are only paying full price for the ones that already proved something.
Recruit for volume, not for names. An open collaboration plan with a competitive commission and samples available can generate fifty to a hundred creator videos a week without direct outreach, and most of those will underperform — which is fine, because the plan is to screen them and amplify the handful that clear. Pair it with a targeted tier: a base rate available to anyone, and a higher rate offered directly to the specific creators whose content has already converted for your category. Blanket rates across all affiliates are the common mistake; they overpay for volume you were going to discard anyway.
How DAMI produces, measures and recruits at the pace this needs
DAMI brings an 8-million-plus creator database, competitor and competitor-store creator search, automated outreach tasks, AI-assisted outreach, AI video creation, sample management, campaign management, social media management, store analytics, and multi-store team dashboards into one place. The constraint this creates is production speed, and three DAMI functions address it directly:
Keep the pipeline full of new variants
Use AI video creation to turn a winning creator video into additional variants for testing, so the weekly cadence in step three does not depend on how many creators replied this week. The point is fifteen to twenty distinct concepts in the pipeline, not fifteen edits of one concept.
Recruit at the volume the screening needs
Run an open plan and a targeted tier together from campaign management, with sample dispatch handled inside the same record, so the fifty-to-a-hundred-a-week pipeline in step four is a setting rather than a manual recruitment drive.
Read cost per outcome, not cost per click
Use store analytics to hold every test against contribution margin after commission, shipping and refunds, so the scaling decision in step three is made on the number that actually decides whether the spend was worth it.
Frequently asked questions
My CPC is already near $1.50. Is Spark Ads really going to change that?
It changes what you pay for the same reach rather than changing the market price. Spark Ads typically run 30 to 50 per cent below equivalent brand-made in-feed CPC, in the $0.20 to $0.50 band, because existing engagement is read as a quality signal. It will not rescue a product whose margin cannot support any paid acquisition at all.
How long before I know whether a piece is worth amplifying?
Two days of organic life, then four to five days of small paid spend. Reading earlier than 48 hours organic mostly measures when the creator happened to post; reading earlier than four days paid mostly measures the learning phase.
Should I stop running product shopping ads entirely?
No. They have a role when a listing is already converting well and you want incremental volume at a known cost. The argument is about which comes first: prove the creative, then buy placement, rather than buying placement and hoping the creative holds up.
Is this just a way to spend less?
No, and that version of it fails. If you cut budget without changing what you amplify, you buy less of the same expensive thing. The saving comes from moving spend onto content that already earned attention for free, which is a different purchase at a different price.
Can I run the screening manually at first?
Yes, and you should — three signals, three to five pieces and a week of readings fit in a sheet, and running one cycle by hand teaches you what your own thresholds actually are. It stops being the right setup when you are screening dozens of pieces a week across several products, or when production, recruitment and measurement are being tracked in separate files by separate people. Consolidate then, not before.
Conclusion
Distribution is the one input you cannot buy back at last year's price, and everything above is about getting it from somewhere the price did not move. DAMI brings an 8-million-plus creator database, competitor and competitor-store creator search, automated outreach tasks, AI-assisted outreach, AI video creation, sample management, campaign management, social media management, store analytics, and multi-store team dashboards into one place.
Pull your last seven days of ad spend and write down CPC alongside contribution margin per order after commission, shipping and refunds. If CPC has risen while that margin has not, the problem is what you are amplifying rather than how much you are spending, and you now know where to start. Ten minutes.
Start with DAMI and start the next cycle from traffic you did not have to bid for.
