Bartley Studio Get your price

Meta ads and agencies

We spend $60k a month at 1.4 ROAS and have been burnt by agencies and media buyers. How do we choose the next one?

By Dan BartleyUpdated 3 minute read

The short answer

Start by working out whether 1.4 is the problem: compare it with your breakeven ROAS and your blended MER, because at some margins 1.4 is profitable and at others no agency can fix it. Then choose on evidence: a written diagnosis of your account, a creative plan with a monthly number, reporting against your margin, your ownership of every asset and a short first term.

Is a 1.4 ROAS actually the problem?

Only if it sits below your breakeven ROAS, which comes from your margin rather than a benchmark. Divide 1 by your contribution margin after product cost, freight, payment fees and GST: at a 50% margin breakeven is 2.0, and at 75% it is about 1.33.

Then check the numbers Meta cannot see. A 1.4 platform ROAS can sit alongside a healthy business if your store's blended figures are strong and new customers keep buying.

  • Blended MER: total store revenue divided by total ad spend, from Shopify rather than Meta
  • New-customer ROAS or CAC, because returning buyers flatter the platform number
  • Repeat purchase rate: whether a first order near breakeven pays back on the second
  • Contribution after ads: gross margin minus ad spend, in dollars, month by month

Why do agency relationships keep failing at this spend?

Usually because the job was scoped as media buying when the constraint was creative, offer or margin. At $60k a month an account needs a steady flow of new concepts, and a buyer who can only move budgets will plateau, whoever they are.

Look back at the last two relationships and ask what changed in the account each month. If the answer is mostly budgets, bids and audiences, the next partner needs to bring creative volume, or you need to supply it.

How should we evaluate the next agency or media buyer?

Ask for evidence before you sign: a written diagnosis of your account, a plan for the first 90 days and the creative they would run. A partner who has read your account answers with your numbers, not a deck.

What to ask, and what the answers tell you
AskA good answerA warning sign
What is wrong with our account?A specific diagnosis with numbers, after reading itA generic pitch about scaling
How many new concepts a month, and who makes them?A number and a named team"We will use your existing assets"
What target will you report against?Breakeven ROAS, MER and new-customer CACPlatform ROAS alone
Who owns the ad account and the creative?You do, with partner access for themThe account sits in their portfolio
What happens if it is not working by month three?A short first term and a clean exitA 12-month lock-in with exit fees

Should we start with a short paid term?

Yes, a short, paid first term is a fair middle ground. Make it long enough for several rounds of creative tests, which for most brands means two to three months, and short enough that neither side is trapped.

Agree before day one what the first term should prove: the number of concepts tested, what was learned, and where MER and new-customer CAC should sit by the end.

What should we fix before anyone new starts?

Anything outside the ad account that caps results: margin, offer, stock depth and the product page. A new team inherits all of them, and naming them up front stops the next relationship failing for the same reasons.

  • A margin that leaves room for ads after freight and fees
  • An offer worth clicking, compared honestly with competitors' offers
  • Enough stock in the products the ads will push
  • Product pages that convert the traffic you already have
  • Tracking that matches Meta's purchases to Shopify orders

Other questions people ask about this

Should we hire an agency that guarantees ROAS?

Treat a guarantee as a warning sign. ROAS depends on your offer, margin, stock and product page as well as the ads, so a promised number is either a sales line or a sign it will be measured on the most generous attribution setting.

Read the full answer
How long before we judge the new team?

Judge the first month on what they tested and learned, and results on two to three months of tests. If they cannot explain each change and what it taught them, you will know sooner.

Should we bring it in-house instead?

At this spend, in-house is realistic. The deciding factor is creative supply: whichever buyer you choose still needs new concepts every month.

Read the full answer

Still have a question? Ask us

Dan Bartley

Founder, Bartley Studio

Dan runs Bartley Studio. He has spent eight years in design, creative direction and paid ads, and now directs the imagery, video and Meta ad accounts the studio delivers for e-commerce brands in Australia and New Zealand.

Get your price in about a minute

Price my whole catalogue

A short quiz gives you a tailored price before you speak to anyone.