Case study
One channel, one monthly newsletter and platform numbers that did not match the store. Over six months we fixed the measurement, built a weekly creative test cadence, and turned email and SMS into flows that earn on their own.
Named only with client approval Figures marked are illustrative Reply within one business day
Attribution · this quarter
GA4 · server-side312
Booked calls
3.1x
ROAS
−31%
CPL
A founder-led direct-to-consumer outdoor apparel brand shipping across Canada depended entirely on paid social, which stopped paying for itself after a platform pricing change.OlDevs set up server-side conversion tracking reconciled against order data, moved creative to a weekly testing cadence of customer-shot video, founder explainers and comparison formats, and rebuilt email and SMS into lifecycle flows covering welcome, browse and cart recovery, post-purchase care and winback. Product page and checkout tests ran alongside, changing the economics of every paid click. Blended return on ad spend reached 3.1, up from 1.6, with 27% of monthly revenue now coming from email and SMS flows.
Key facts
The challenge
Paid social was the brand's only growth channel, so a platform pricing change was enough to push the whole business under water. Email amounted to one monthly newsletter sent to the entire list, and most creative was reshot product photography that looked like every other outdoor brand in the feed. Worst of all, the founder was making budget calls from platform-reported figures that did not match the store's own order data.
Every new customer arrived through paid social, so a change in auction pricing hit revenue immediately with nothing to absorb it. There was no owned channel producing sales, and no way to slow spend for a month without watching orders fall off at the same rate.
Platform-reported conversions and the store's order data disagreed. The founder was choosing which campaigns to fund from figures nobody could tie back to shipped orders, so good campaigns were cut and weak ones kept running.
Ad creative was reshot studio product photography, indistinguishable in a feed from every competing outdoor brand. Production was slow per asset, so there were never enough variants to learn from and losing ads stayed live far longer than they earned.
One send a month went to the whole list regardless of what anyone had browsed, bought or abandoned. Customers who added to cart and left heard nothing, buyers got no care instructions, and lapsed customers were never asked to return. The list earned almost nothing.
What we built
We set up server-side conversion tracking and reconciled it against order data, so spend decisions rested on what actually shipped rather than what the platform claimed. Creative moved to a weekly testing cadence of short customer-shot video, founder explainers and comparison formats, with losers cut on a fixed schedule instead of on instinct. Email and SMS were rebuilt into lifecycle flows so revenue no longer depended on the day's ad auction, and product page and checkout tests raised conversion enough to change the economics of every paid click.
Conversions sent server side and checked against the store's own order data, with the gap explained rather than ignored. Budget decisions moved onto figures that match what shipped, which changed which campaigns survived the next review.
Short customer-shot video, founder explainers and comparison formats produced and launched on a weekly rhythm, with a fixed rule for cutting losers. Volume replaced polish, so the account learns continuously rather than waiting on the next studio shoot.
Welcome, browse abandonment, cart recovery, post-purchase care and winback built as flows rather than sends, triggered by what each customer did. Revenue from the list now arrives every day of the month instead of on the day a newsletter goes out.
A sequence of tests on the product page and checkout: sizing and fit clarity, shipping expectations, and fewer steps between add to cart and payment. Each win improves the return on every click the brand has already paid for.
Process
Weeks 1 to 3: Measurement before spend
Server-side tracking went in first and was reconciled against store orders, so every later decision could be argued from shipped revenue rather than platform reporting.
Weeks 2 to 6: Creative system
We set the weekly production and testing rhythm, sourced customer-shot footage, filmed founder explainers, and agreed the fixed rule for cutting an ad that was not earning.
Weeks 4 to 12: Lifecycle build
Flows were built and launched in order of expected return: cart recovery first, then welcome and browse abandonment, then post-purchase care and winback sequences.
Weeks 8 to 20: On-site testing
Product page and checkout tests ran in sequence against real orders, with each result documented so the founder could see what changed and why conversion moved.
Weeks 20 to 26: Handover and cadence
We documented the flows, tracking setup and creative rules, then handed over the weekly routine so the team can run the cadence and read the numbers without us.
Stack
Storefront
Paid social
Measurement
Email and SMS
Creative
3.1
Blended return on ad spend, up from 1.6
27%
Of monthly revenue now from email and SMS flows
+19%
Site-wide conversion rate after product-page and checkout tests
Outcomes
Blended return on ad spend reached 3.1, up from 1.6, driven by honest measurement, a weekly creative cadence and a checkout that converts more of the traffic already paid for.
27% of monthly revenue now comes from email and SMS flows. That share arrives whether or not the platform changes its pricing, which is what makes the rest of the plan survivable.
Site-wide conversion rate rose 19% after the product-page and checkout tests. The same traffic, at the same cost, produces more orders, which is why the paid numbers moved as far as they did.
Spend decisions, inventory buys and forecasts now rest on figures that reconcile with shipped orders. The weekly rhythm of test, cut and report is documented, so the team runs it without waiting on an agency.
In their words
I stopped waking up to check the ad account. The flows run, the numbers reconcile, and I can plan inventory against them.
Founder, direct-to-consumer apparel brand
E-commerce & retail
FAQ
Browser-based tracking loses events to consent choices, ad blockers and privacy defaults, and platforms model the gap in their own favour. Sending conversions server side and reconciling them against order data gives you one set of numbers you can plan inventory against, even when the two sources still differ.
Enough that losing an ad costs nothing. This brand moved to a weekly cadence of short customer-shot video, founder explainers and comparison formats, with a fixed rule for cutting underperformers. Volume matters more than polish here, because the account only learns from variants it has actually served.
Cart and checkout recovery, almost always, because the intent is already there and the audience is small enough to launch quickly. Welcome and browse abandonment follow, then post-purchase care and winback. Built in that order, the flows start returning revenue while the rest of the programme is still being written.
You get the option, which is the point. When a meaningful share of monthly revenue comes from owned channels, a pricing change or a weak creative month is no longer an emergency. Most brands keep spending, but they spend from choice and at a return they can measure.
Yes. Clients own all code, designs, accounts and IP, including the ad accounts, tracking setup, lifecycle flows and test documentation. We hand over the weekly routine so your team can run it, and we stay involved only where that is useful. Request a quote and we will reply within one business day.
Let’s connect
Tell us what is not working. We’ll reply within one business day with how we would approach it and a tailored quote.