How to Vet a Paid Social Agency on ROAS, Break-Even and New-Customer Growth

Question 1

Should I hire an agency that promises to improve my ROAS?

Accepted Answer

The article advises against hiring the agency that leads with "we improve ROAS" and favors the one that asks permission to lower reported ROAS to buy new customers. According to the article, in documented lift tests roughly 60% of retargeting conversions would have happened anyway and true incremental return often runs 40 to 70% below what the platform dashboard shows. It describes an account spending $20k a month with prospecting at 1.6x that reported 3.4x after shifting 70% of budget to warm audiences, and a brand reporting 4.1 ROAS where a three-week geo holdout showed revenue in paused markets fell by only a third of claimed impact, meaning two-thirds would have happened with ads off.

Question 2

How can reported ROAS rise while total sales stall or fall?

Accepted Answer

The article describes three mechanisms: cutting prospecting that prints at 1.2x while retargeting prints at 5x, which lifts the blended ratio for about two weeks before the retargeting pool starves; pushing about 80% of spend to past visitors and past buyers to report around 4x that business accounts cannot find; and letting platform attribution on a seven-day click plus one-day view window with modeled conversions claim sales, producing splits like 4.2x in the platform versus about 1.8x on net-of-returns revenue across all spend. The article's rule is that the platform number shows what the platform credits itself for, while growth is how many extra dollars showed up.

Question 3

Why does a percentage-of-spend fee work against good budget decisions?

Accepted Answer

The article states agencies often charge 10 to 20% with a minimum, so finding $2,000 in wasted spend and switching it off gives a typical 15% deal a $300 monthly pay cut for doing the job well. According to the article, that structure plus a ROAS-first promise favors keeping spend on audiences with the safest reported ratio and postponing decisions to spend more on strangers or spend less when saturated. The article states groas runs management for a flat $1,499 a month with no cut of ad spend, so the manager can recommend a smaller budget without taking a smaller fee.

Question 4

How do I set a break-even ROAS and check new-customer growth?

Accepted Answer

The article recommends writing down break-even ROAS as one divided by gross profit margin before setting a target: $100 revenue minus $60 in product, shipping and fees leaves 40% margin, so break-even is 2.5x. It gives skincare at 70% margin with break-even at 1.43x where 2.0x can be profitable, and furniture at 25% margin with break-even at 4.0x where 3.5x loses money. The article also recommends managing new-customer ROAS and CPA against lifetime value with cited triggers of pulling back around 1.7 new-customer ROAS and pushing around 2.1, and checking monthly with blended ROAS defined as total revenue net of returns from the order system divided by total ad spend across platforms.

Question 5

What sales-call questions reveal whether an agency will polish ROAS instead of growing revenue?

Accepted Answer

The article recommends asking for break-even ROAS in the proposal based on margin, shipping, fees and returns, such as 2.5x, and rejecting a 4x promise made without asking margin. It recommends asking what share of revenue will come from new customers and how event coverage will be checked, and what happens to reported ROAS if prospecting spend rises 30% — expecting a frank answer that it may drop for two to three weeks while the pool refills. It recommends asking how the manager will reconcile platform reports with net revenue divided by all-platform spend each month, and what test such as a three-week holdout, geo split or pause-test with a falsifiable prediction will show what would vanish with ads off.

Question 6

When is it right to hold ROAS high instead of buying growth?

Accepted Answer

The article states the one owner who should hold ROAS high is the one who cannot afford to buy growth yet, when cash is tight, margin is thin after returns, and a bad month of prospecting means missing payroll. It gives as arithmetic, not cowardice, that a brand with 25% margins and no repeat purchase has no business paying 1.8x to acquire a stranger. For others, the article argues the agency that names break-even, checks new-customer share, reconciles revenue and asks to spend below today's reported return is preferable even if the report looks worse before the business gets bigger.