Roas Hack

We don’t run ads. We engineer growth.

[Roas Hack @ROAS_HACK]

Most business owners have zero visibility into their unit economics — true CAC, break-even ROAS, or whether their funnel is actually viable.
Our free Unit Economics Calculator shows exactly where you stand (and where the leaks are) in under 60 seconds:
👉 https://t.co/K9buJ2lmw1
Great ads alone won’t win the game. Growth isn’t magic, it’s our strategy.
RT if this hits close to home
Reply with your current ROAS 👇

[Roas Hack @ROAS_HACK]

@CarlWeische @marekols Curious what the actual unlock was. Most fired-then-built-it stories hinge on one constraint that forced brutal focus, usually nailing one offer and one channel before touching anything else. Did he fix the offer first or the funnel first?

[Roas Hack @ROAS_HACK]

Half the time the glitch isn't spend efficiency, it's attribution. Meta over-credits in these windows, modeled conversions inflate, reported ROAS spikes while incremental lift barely moves. Run a geo holdout in a glitch month and most of the lift disappears. The real demand survives the holdout. The glitch doesn't.

[Roas Hack @ROAS_HACK]

"High CAC is fine, the LTV supports it" is the most expensive sentence in DTC.

Sometimes it's true. Usually it's the excuse you use to keep funding a channel that pays back in 14 months while your cash dies in 4.

Be honest about those two numbers. LTV is money you hope to earn, on retention you haven't delivered yet. CAC is money already gone, today, in real pounds.

You can bet present cash on future loyalty if you can fund the wait. Most can't. They find out in month four, when the account's empty and the LTV is still theoretical.

Payback in days tells you how fast you can scale. LTV tells you a bedtime story.

What's your real payback window, and does your bank balance back it up?

[Roas Hack @ROAS_HACK]

The tell is when the angle is about the trend, not the customer. ai, glp-1, subscription all convert only if the product already earns them. Bolt a macro narrative onto one that doesn't and you win the swipe, lose the checkout. The boring angle that names the actual pain usually beats all three.

[Roas Hack @ROAS_HACK]

The thing I'd rule out first: a 7-10 day post-click drop hitting a lot of accounts at once is also the exact signature of a signal or attribution shift, not just audience quality. Quick check before you rebuild around it. Does the drop show up in reconciled backend revenue and new-customer AOV, or only in Meta-reported numbers? If the backend holds, it's measurement and you fix the pixel/CAPI. If the backend confirms it, it's real and your levers are the right call.

[Roas Hack @ROAS_HACK]

Your bestseller is probably losing you money. You just can't see it under the volume.

Bestsellers usually win on price. Steep discount, high returns, heavy shipping. The revenue line looks incredible, so nobody audits the margin underneath it.

Do this once. Take each SKU's real selling price and subtract:

what the product costs you
shipping and fulfillment
payment fees
returns on that exact SKU
the discount it actually sells at, not list price
Re-rank the catalog by what's left. Not revenue. Contribution.
The list reorders, sometimes violently. I've seen the hero product run negative while a boring SKU nobody mentions quietly funded the whole business.
When did you last rank SKUs by contribution instead of sales?

[Roas Hack @ROAS_HACK]

You proxy it. The clean number, CM per order under the new index, doesn't exist in most stacks. So you watch what moves first: nCAC against blended CAC, AOV on new customers only, return rate lagged a month, repeat rate on that cohort. CPA fell and those hold, it repriced. They slip, it deteriorated and the average hid it.

[Roas Hack @ROAS_HACK]

@iamshackelford Strong, the offer stack is underrated. One caution on 'initial conversion isn't profitable, LTV makes it profitable': only if you can fund the payback gap. Adding 3-5 offers that each lose up front can drain cash before LTV lands. Track payback per offer or you scale the bleed.

[Roas Hack @ROAS_HACK]

@binghott @codyplof @NotZainAgain It's a double tax. They take credit for a sale that would've mostly happened anyway, so you pay a commission plus a discount on demand you already created. And you lose the owned relationship on top. Only real test is incrementality: kill the coupon and see what disappears.

[Roas Hack @ROAS_HACK]

In affiliate marketing, nobody pays you for a great campaign. They pay you when the math works. That's the whole job.

No brand budget to hide in. No awareness excuse. If the offer flopped and the numbers didn't clear, you ate it that day, with your own money.

You learn one habit fast. Stop admiring the campaign. Start asking if more came back than went out.

DTC dodged that question for a decade because cheap clicks covered a lot of bad math. Clicks aren't cheap now. A lot of brands are about to learn their unit economics never worked, they were just subsidized.

What's a metric you used to trust and now ignore?

[Roas Hack @ROAS_HACK]

@Seanfrank The blue collar framing is right, but the 1-2 who take home the 10M usually didn't win on CPM. They won on contribution margin and payback. CPM hacking is the visible grind. The quiet margin discipline decides who keeps the money, not who makes the revenue.

[Roas Hack @ROAS_HACK]

@BillDA The trap we see even with owners who read their P&L: they confuse gross margin with contribution margin. Looks fine on paper, but load in shipping, fees, returns and discounts per order and a chunk of 'winners' are underwater. Literate, and still blind on the number that matters.

[Roas Hack @ROAS_HACK]

@oyemahi_ @andrewjfaris to make sure new ad get spend do 1 1 1 setup

[Roas Hack @ROAS_HACK]

@codyplof The falling CPC is the tell it's not only pixel. Measurement loss would tank reported CVR but not your CPC. Both dropping together points to delivery: Advantage+ pushing impressions into cheaper, lower-intent inventory. Lower CPC, lower CVR, CPA roughly flat.

[Roas Hack @ROAS_HACK]

Your ad platform reports one business. Your bank account reports another. Only one of them signs your paychecks.

Meta says 4.1x. You reconcile against real deposits, after returns, after discount codes, after the sales it claimed but never caused, and you're sitting at 2.3x.

Both are real. One is marketing. One is money.

The platform is built to look good enough that you keep spending. Your P&L isn't built to flatter anyone. It just tells the truth a month late, when it's expensive to hear.

Reconcile one month of platform revenue against your actual deposits. The gap is the lie you've been scaling.

Have you ever run that reconciliation? What was the gap?

[Roas Hack @ROAS_HACK]

TikTok CPMs jumped 58% this year. That's not the thing killing you.Everyone's panicking about the jump from $6 to $10. Meanwhile their creative dies in about 72 hours and they refresh it once a month. You're paying premium prices to keep running dead ads.The CPM isn't the problem. Your refresh cadence is.

[Roas Hack @ROAS_HACK]

TikTok CPMs jumped 58% this year. That's not the thing killing you.Everyone's panicking about the jump from $6 to $10. Meanwhile their creative dies in about 72 hours and they refresh it once a month. You're paying premium prices to keep running dead ads.The CPM isn't the problem. Your refresh cadence is.

[Roas Hack @ROAS_HACK]

I changed my mind about killing ads fast.

I used to cut anything under target within 3 days. Felt disciplined. I was actually murdering ads before they had the data to prove anything, then wondering why nothing scaled.

Now I'd rather give a campaign enough spend to be real than kill it on one bad Tuesday. Patient with the data, ruthless with the conclusion.

Speed felt like rigor. It was just chasing noise.