Ecommerce case study

3.4× return on ad spend, in two quarters.

Go Swag was running paid through two agencies and an in-house hire, and the blended return had sat flat for three quarters. We took paid acquisition onto one team, with one set of numbers, and turned the flat return around on a flat budget.

Why it mattersA 1.6× blended return usually means you throttle spend. Getting to 3.4× on a flat budget is what let them scale instead of pull back.

Go Swag · Ecommerce · Meta and Google Ads, plus email flows · Two quarters to the turn
The whole story is on this page
The short version

Three teams, and nobody owned the whole picture.

Go Swag was running paid across two agencies and a recent in-house hire when we plugged in. Nobody owned the whole picture, the tracking didn't agree with itself, and blended return had been flat for three quarters while spend kept creeping up. The goal: one team, one set of numbers, and a return that justified scaling instead of throttling.

What they had going for them

  • A product that sold. The sales were real, whoever claimed them.
  • Budget already committed to paid, and it stayed flat through the turn.
  • Email flows already in place, carrying 9% of revenue.
  • A clear goal from day one: scale, if the return could justify it.

What was quietly costing them

  • Two agencies and an in-house hire, with nobody owning the whole picture
  • Tracking that didn't agree with itself, each platform claiming the same sales
  • A blended return flat at 1.6x for three quarters while spend kept creeping up
  • Branded-search overspend claiming credit for sales it didn't create
The mess we walked into

Three dashboards. Three versions of the truth.

The 30-Day Plug-In audit found three versions of the truth, with each platform claiming credit for the same sales. Every dashboard graded its own homework, and every dashboard passed.

Agency one's dashboardclaimed the sale
Agency two's dashboardclaimed the same sale
The bank accountsaw one sale
one dashboard, one set of numbers
Blended return3.4xfrom 1.6x
Days to payback44from 71
Cost per customer−31%same flat budget
Email share of revenue24%from 9%

What replaced the three dashboards: one view, one set of numbers, every spend decision made from the same math. A recreation of the view they run on now.

One set of numbers, before and after
Blended return, then
1.6x
Blended return, now
3.4x
Days to payback, then
71 days
Days to payback, now
44 days
Cost per customer, then
baseline
Cost per customer, now
down 31%

For days to payback and cost per customer, the shorter bar is the better one. Bar lengths are scaled within each pair.

What we did about it

One team, one set of numbers. Meta and Google run together, tracking rebuilt so every spend decision used the same math, and email put to work carrying revenue that paid had been getting charged for.

The check-up

Then and now.

Every engagement starts with the same audit we'd run on your business. Here's how Go Swag's paid acquisition graded when we plugged in, and how it grades today.

What we checkedWhere it stood when we plugged inThen → now
Who owned paid
Two agencies and a recent in-house hire; nobody owned the whole picture
FA

Paid acquisition moved onto one team, with Meta and Google run together. One owner for the whole picture, so no dashboard gets to grade its own homework anymore.

Tracking
Didn't agree with itself; each platform claimed credit for the same sales
FA

Rebuilt so every spend decision used one set of numbers. Three versions of the truth became one, and the branded-search overspend that claimed credit for sales it didn't create had nowhere left to hide.

Blended return
Flat at 1.6x for three quarters while spend kept creeping up
DA

3.4x blended within two quarters, on a flat budget. The return that turned the conversation from throttling spend to scaling it.

Email
Carrying 9% of revenue while paid got charged for the rest
CB+

The flows were reworked to carry more of the revenue. Email now carries 24%, and the repeat rate climbed from 22% to 29%.

Where the budget went
Campaigns that looked busy but never paid back, plus branded-search overspend
DA

The busywork campaigns got cut. Cost per customer came down 31%, and payback tightened from 71 days to 44.

Tap any line for the detail →

Grades are how we score a marketing check-up. Every engagement is different; results shown are not a guarantee. How we present results

What we ran

Same three rituals, pointed at paid.

Nothing exotic. The plan consolidated Meta and Google under one team, rebuilt tracking so every spend decision used one set of numbers, and put email to work carrying revenue that paid had been getting charged for. It all started with the same 30-Day Plug-In audit, and it is the kind of work we run for ecommerce brands.

Days 1 to 30

The 30-Day Plug-In

  • Audit, access, and the baseline
  • Found three versions of the truth, one per dashboard
  • Paid consolidated onto one team by day 30, Meta and Google together
Every week since

The Operator Cadence

  • A working session every week, a Friday update every week
  • Every spend decision made from one set of numbers
  • Campaigns that looked busy but never paid back got cut in the open
Day 90 and on

The 90-Day Rebuild

  • Busywork campaigns killed for good
  • Email flows reworked to carry more of the revenue
  • Blended return at 3.4x by the end of two quarters, on a flat budget
What we killed

Branded-search overspend that claimed credit for sales it didn't create.

The numbers

What one set of numbers did.

Two quarters on a flat budget. Every figure here comes from the rebuilt tracking, the same set of numbers every spend decision was made from.

1.6x to 3.4xblended return on ad spend, in two quarters on a flat budget
Down 31%cost to acquire a customer over the same stretch
71 to 44 daysto earn a customer's cost back
The customer side of the ledger
Repeat purchase rate, then
22%
Repeat purchase rate, now
29%
Email share of revenue, then
9%
Email share of revenue, now
24%
Two quarters, one set of numbers
Flat budgetsame spend, two quarters running
One teamMeta and Google run together
One truthevery spend decision from the same set of numbers
3.4x
blended return, up from 1.6x. The return that justified scaling instead of throttling.
9% → 24%
of revenue now carried by email, revenue that paid ads had been getting charged for.

Every engagement is different; results shown are not a guarantee.

The turn

One owner beats three dashboards.

When every platform grades its own homework, everyone passes and nothing improves. One team with one set of numbers fixed the math.

3.4x

Blended return on ad spend, up from 1.6x in two quarters. Same budget, one team, one set of numbers.

31%

Off the cost of acquiring a customer, once the busywork campaigns and the branded-search overspend were cut.

24%

Of revenue now carried by email, up from 9%. Paid stopped getting charged for sales it didn't create.

44 days

To earn a customer's cost back, down from 71. The number that turned “should we cut spend?” into “where do we scale next?”

From the founder

“The first month I just wanted to see them not make it worse. A quarter in we were up, and I stopped opening the ad account at midnight.”

Founder, Go Swag

Every engagement is different; results shown are not a guarantee. How we present results

More of the work: Manako Labs and Kubera Health

Want results like this?

Curious what your own funnel could do? Book a free analysis, or see more of the work.