--- title: Creative economics — the median ad loses money, so cut the cost per ad (CTC's 1.3M-ad study) slug: creative-cost-per-ad-economics summary: CTC's analysis of 1.339M Meta ads across 158 brands — the median ad is worth −$3.20 in 60-day contribution, the average +$507, and the top 5% of ads produce 73% of all value — so you can't change the odds, only the price of each "ticket"; commission-only creator programs (Tribe, Refunnel) that pay only for winners let The Normal Brand triple ad volume and double spend at the same incremental ROAS. domain: marketing tags: [creative-volume, creative-economics, meta-ads, ugc, creator-programs, whitelisting, incrementality, ecommerce] related: [bfcm-ad-creative-volume, meta-ads-creative-testing-engine, ugc-creator-sourcing-for-ecommerce-ads, partnership-ads-and-creator-selection, ad-creative-volume-and-hit-rate, paying-above-first-order-cac, marketing-moments-calendar, creative-demand-formula-carry-rate] sources: [yt:bqJElruBvjE, yt:Z9rcSM01fv8] updated: 2026-10-24 --- # Creative economics: the median ad loses money, so cut the cost per ad ## TL;DR - **Score each ad by 60-day contribution:** conversion value × incrementality factor × 60-day LTV × gross margin − ad spend. - **Across 1.339M ads (158 brands, 2 years):** the median ad is worth **−$3.20**, and this holds at 133 of 158 brands. The average is **+$507**, because value is extremely right-skewed. - **Top 1% of ads = 41% of value. Next 4% = 32%. Top 5% = 73%.** Meta caps the downside by starving losers for free. - Taylor Holiday says no brand he's seen can reliably change these odds. **What you control is the cost of each "ticket".** At ~$300 an ad and one $10K winner per ~80 launches, paid-up-front production doesn't pencil. - **Invert the creative cash cycle:** commission-only creator programs (Tribe, Refunnel, TikTok Shop affiliates) mean you pay only when an ad sells. - **The Normal Brand:** ~380 → ~1,200 new ads a month at zero production cost, spend +100–140% year on year at the same incremental ROAS. The median ad didn't improve. Winners above $10K went from 5 to 16. ## Details **The stuck brand.** The Normal Brand (menswear, mediocre LTV) stalled in Q1 2026. CTC launched 1,199 ads in the quarter (400, 240, 524 by month), and every push for scale broke first-order profitability. Production cost rose with volume because every ad was paid up front.[^1] On the 60-day contribution metric, its median ad was worth about −$2 and its average $24. 85% of ads lost money, and the typical ad cost more than $24 to make.[^2][^3] **The 1.3M-ad study.** The median ad was −$3.20 (negative at 133 of 158 brands) and the average ~$507. Of the ads, 631K lost money, 246K made money and 11K made "big money".[^4][^5] Meta "kills your losers for free": spend on losers is capped, while upside is unbounded.[^6] The top 1% of ads produced 41% of value and the next 4% another 32%.[^7] Taylor's claim is that you can't alter the odds of this lottery, only the price of the ticket. Creative production cost is "the hidden number destroying your ad economics".[^8][^9] **The ticket-price maths.** Assume one $10K winner per ~80 launches and ~$300 per video ad. Taylor puts the cost per winner at ~$26,400, so "it doesn't pencil".[^10] The fix is to make the channel that produces the asset carry its cost, like consignment.[^11][^12] - *Old cycle:* pay for production → pay Meta → maybe get cash back 90 days later. - *Inverted cycle:* seed product → creator submits a video → you whitelist it as a Meta ad → it sells → you pay the creator a commission on attributed sales.[^12] He expects the arbitrage to fade as creators demand a "terminal price of creative" that shares the risk fairly.[^13] **The Normal Brand after switching to Tribe (Apr–Jun 2026).** - Spend rose +126%, +140% and +100% year on year in April, May and June, at the same ROAS and the same *incremental* ROAS.[^14][^15] - New ads went from ~380 to ~1,200 a month. In June, Tribe ads were 78% of live ads and 81.5% of spend.[^14][^15] - The average ad became 5.4x more profitable but still under $300. The economics only work because production cost went to zero.[^16] - Median 60-day spend per ad didn't move ($920 → $1,053 → $924 for Tribe ads), but the 99th percentile rose from $3,300 to $6,800. Ads clearing $10K went from 5 to 16, and Tribe made 12 of them.[^16][^17][^18] - His conclusion: "the ceiling wasn't demand, it was creative supply". Hold your quality bar constant and scale volume against it.[^15][^18] **Curated creator networks (Joy Sharma, CTC).** Refunnel had one-click whitelisting via Meta's API and ~50,000 whitelisted creators. CTC asked it to pull every creator more than one standard deviation above the mean (~3,000 people) into a community. It offers flat fees rather than a percentage of spend or revenue, and it handles tax forms (1099s) that DIY creator programs pile up.[^19][^20][^21] (Offered free to CTC clients, which is a commercial tie.) ## Disagreements / caveats - **Hit rate:** Smart Marketer's SOP targets 20–30% "winners" in discovery tests ([[meta-ads-creative-testing-engine]]). CTC measures *contribution value* and finds most ads lose money. Different definitions, but a big gap in optimism. - **Quality vs volume:** Taylor concedes better ads improve the economics further but says median quality didn't change. Quality-first voices disagree. See [[ad-creative-volume-and-hit-rate]] for the app-UA version of this debate. - CTC sells creator-program management and has commercial ties to Refunnel. Tribe was named from the stage. These are self-reported client results. ## See also - [[bfcm-ad-creative-volume]] — the same "make more ads" logic for Black Friday - [[ugc-creator-sourcing-for-ecommerce-ads]] — the paid-up-front creator model this replaces - [[partnership-ads-and-creator-selection]] [^1]: yt:bqJElruBvjE @ 16:55 — Taylor Holiday [^2]: yt:bqJElruBvjE @ 17:42 — Taylor Holiday [^3]: yt:bqJElruBvjE @ 18:29 — Taylor Holiday [^4]: yt:bqJElruBvjE @ 19:15 — Taylor Holiday [^5]: yt:bqJElruBvjE @ 20:01 — Taylor Holiday [^6]: yt:bqJElruBvjE @ 20:53 — Taylor Holiday [^7]: yt:bqJElruBvjE @ 21:43 — Taylor Holiday [^8]: yt:bqJElruBvjE @ 22:32 — Taylor Holiday [^9]: yt:bqJElruBvjE @ 23:22 — Taylor Holiday [^10]: yt:bqJElruBvjE @ 24:11 — Taylor Holiday [^11]: yt:bqJElruBvjE @ 24:57 — Taylor Holiday [^12]: yt:bqJElruBvjE @ 25:44 — Taylor Holiday [^13]: yt:bqJElruBvjE @ 26:33 — Taylor Holiday [^14]: yt:bqJElruBvjE @ 27:19 — Taylor Holiday [^15]: yt:bqJElruBvjE @ 28:09 — Taylor Holiday [^16]: yt:bqJElruBvjE @ 28:57 — Taylor Holiday [^17]: yt:bqJElruBvjE @ 29:44 — Taylor Holiday [^18]: yt:bqJElruBvjE @ 30:30 — Taylor Holiday [^19]: yt:Z9rcSM01fv8 @ 20:32 — Joy Sharma [^20]: yt:Z9rcSM01fv8 @ 21:17 — Joy Sharma [^21]: yt:Z9rcSM01fv8 @ 22:03 — Joy Sharma