All answers

The Prevailing
Answers for
Business model:
Business model, pick up to two

Where should we set cost caps or ROAS floors relative to current results?

The PrevailingPublished Oct 12, 2026

Our answer

Anchor caps to the cost you can afford, not your average. Set a cost per result goal about a third below your limit, and a ROAS goal close to your target.

Confidence
Medium
Strongest proof
4 Meta rules
6 datasets, 10 practitioners, 2 opinions or claims
Sources
8 sources
22 statements

New here? Take the 1-minute tour

Why this is the answer

3 reasons and our judgment

Cost per result goals overshoot their target, while ROAS goals land close. Cost per result goals delivered 140% of target cost per purchase, about 40% over, while ROAS goals delivered 96.5% of target (Common Thread Collective, or CTC, from 253 accounts).

Missed targets are common enough that Common Thread Collective corrects for them. It applies correction factors of 154% for cost goals and 95.9% for ROAS goals, and found 75% of accounts on cost per purchase goals ran more than 5% over target.

Bid the threshold you can accept, not your average. Andrew Faris does this, and under Incremental Attribution he sets the cap a little low.

Our judgment

We start from Meta's suggested goal, 10 to 20% above average cost, only when your average cost sits comfortably inside your limit, and otherwise anchor to the limit. That reconciles Meta's rule of starting 10 to 20% above average with CTC's overshoot finding and Faris's threshold bidding. For a catalog with a wide range of order values, we judge each cap as a ROAS per product group, extending Faris's single-product example of a $77 cap on about a $97 order.

Our judgment, built from Meta, Common Thread Collective and Andrew Faris: our reading of these sources, not a finding from any one of them.

What the sources say

22 statements from 8 sources
  • MetaUndatedSupports

    Meta says to start a ROAS goal 10 to 20% below your average ROAS (a typical 1.0 means a goal of 0.8 to 0.9), judge results on weekly averages, and adjust 5 to 10% at a time.

    “For example, if you typically see a ROAS of 1, try setting your goal around 0.8-0.9 range.”

    Read it on facebook.com

  • MetaUndatedDisagrees

    Meta says to set a cost per result goal 10-20% above your average cost per result, run it 3 weeks before scaling, wait 7 days between changes and move it 5-10% at a time.

    Read it on facebook.com

  • MetaUndatedDisagrees

    Meta's bid cap example: a brand that loses money above a $40 cost per purchase sets a $50 bid cap, because the cap limits the auction bid, not the cost per result.

    “the cap sits above the target cost because it limits the auction bid, not the result cost.”

    Read it on facebook.com

  • MetaUndatedCaveat

    Meta's bid strategy documentation says Highest Volume spends the full budget, while cost cap, bid cap and minimum ROAS can under-deliver. Cost cap limits are not guaranteed.

    “adherence to cost cap limits is not guaranteed”

    Read it on developers.facebook.com

  • Marin IstvanicOct 27, 2026Supports

    Marin Istvanic reads bid caps by cost and spend: on target but underspending, raise the bid. On target at full spend, raise budget. Over target at full spend, lower the bid. Over and underspending, new ads.

    “Good CPA + not spending full budget → increase the bid. It's too conservative.”

    Read it on x.com

  • Cody WittickOct 15, 2026Supports

    Kynship's Cody Wittick sets spend from the P&L: what a new customer can cost becomes the aMER floor, which sets the cost caps that decide how much the account may spend.

    “The aMER floor sets the cost caps. The cost caps decide how much the account is allowed to spend, and they flex when demand does.”

    Read it on x.com

  • Andrew FarisSep 8, 2026Supports

    Andrew Faris's sale playbook: launch with a loose cost cap under a hard budget ceiling; once on target, raise budget 50% and cut the cap 20%; the day after, cut the cap about 50%. The numbers are illustrative.

    Read it on youtube.com

  • Antonio VentreSep 7, 2026Caveat

    Antonio Ventre's safe bid cap option sets the bid at your target cost per acquisition, which he says works best early on with a clear winning ad, strong engagement and low frequency.

    “Set the bid at your target CPA: this way, you will be almost 100% certain that if you get purchases, they will be under your CPA target”

    Read it on youtube.com

  • Common Thread CollectiveAug 25, 2026Supports

    CTC analyzed about 320 Meta accounts and nearly $1.5B in spend: minimum ROAS landed within several percent of target across the portfolio, while cost per result goals hit target less than half the time.

    “the cost per result goal only achieved its target less than half of the time”

    Read it on commonthreadco.com

  • Common Thread CollectiveAug 25, 2026Caveat

    Common Thread Collective says a blended ROAS target is not the same as the on-platform target: with a 2.0 business need and Meta's 7-day click incrementality factor of 1.2, the on-platform target is about 1.8.

    “Your blended ROAS target (what the business needs) is not the same as your on-platform bid target.”

    Read it on commonthreadco.com

  • Common Thread CollectiveAug 6, 2026Supports

    CTC's study of 253 Meta accounts and $200M of spend found minimum ROAS averaged 96.5% of target but 67% of accounts missed by over 5%; cost-per-result goals landed at 140% and bid caps at 123%.

    “Bid Cap Delivered 123% of Target and Is Now CTC's Default”

    Read it on commonthreadco.com

  • Common Thread CollectiveAug 6, 2026Supports

    CTC applies correction factors to bids: about 95.9% for minimum ROAS and 154% for cost per result goals, meaning a cost cap set roughly a third below the CPA it actually wants.

    Read it on youtube.com

  • Common Thread CollectiveAug 6, 2026Supports

    Common Thread Collective, across 253 accounts, says account-level misses are common: 75% of accounts on CPA goals ran more than 5% above target, and 67% on minimum ROAS ran more than 5% below.

    Read it on youtube.com

  • Andrew FarisAug 6, 2026Supports

    Andrew Faris waits about 50 purchases before changing a manual bid, and says the bid only needs to be consistent, not accurate, since you keep adjusting it toward the target.

    Read it on youtube.com

  • Common Thread CollectiveAug 6, 2026Supports

    Taylor Holiday says Meta's cost per result goals ran over target in Common Thread Collective's data: on 7-day click attribution, results came in at 140% of target cost.

    “So seven eight click was at 140% of target.”

    Read it on podcasters.spotify.com

  • Common Thread CollectiveAug 6, 2026Supports

    Taylor Holiday of Common Thread Collective says across a broad dataset, Meta's minimum ROAS bidding delivered against the set target 97% of the time.

    “Meta's unbelievable capacity to deliver this outcome across a broad enough data set... 97% actual to target.”

    Read it on podcasters.spotify.com

  • Andrew FarisJul 15, 2026Supports

    Andrew Faris says to set a cost cap at the highest cost you can accept, not at your average cost.

    Read it on youtube.com

  • Andrew FarisJul 15, 2026Supports

    Andrew Faris sets cost caps under Incremental Attribution a little low, because a cost cap averages out and will buy some purchases above his target CAC.

    Read it on youtube.com

  • Andrew FarisJul 15, 2026Supports

    Pushes back on bidding to an average: Andrew Faris says that with Meta cost cap campaigns, advertisers should bid on a threshold rather than an average.

    “bidding on a threshold, not on average. This has always been my argument.”

    Read it on youtube.com

  • Jon LoomerJul 2, 2026Caveat

    Jon Loomer cautions that setting bid controls (cost per result goal, bid cap, ROAS goal) too aggressively can starve delivery, because Meta no longer has to spend the full budget.

    “setting bid controls (Cost Per Result Goal, Bid Cap, ROAS Goal) too aggressively can starve delivery, since you’ve removed Meta’s requirement to spend the full budget.”

    Read it on jonloomer.com

  • Andrew FarisJun 25, 2026Supports

    Andrew Faris judges a cost cap as a ROAS (order value divided by cap). On one single-product ad set with a $97 order value he set a $77 cap, and he waits for at least 30, ideally 50 to 100, purchases.

    Read it on youtube.com

  • FlightedJun 5, 2026Disagrees

    Flighted, a Meta agency, says new campaigns without cost history should start on Highest Volume, then add cost or ROAS goals set from the trailing 30 days, not aspirational targets. No data shown.

    “For new campaigns without baseline data, Highest Volume is typically the right starting point”

    Read it on flighted.co

In the full answer

For readers who sign in
  • What to do this week3 steps
  • The numbers behind it3 figures from cited sources
  • Where each source standsA chart placing each source on the question
  • The strongest disagreementThe best case against the answer, from a cited source
  • How sure we areWhat its confidence rests on, and what would change the answer

Read next

More in Bidding and optimization