Between April and August 2026 we cut nothing, fixed the funnel, built a new acquisition channel, doubled the conversion rate, sold 12.5% more orders, and finished the period with 6.1% less revenue than the same window a year earlier.
Most case studies you read stop at the flattering number and go home. This one is more useful than a clean win, because the shape of it is extremely common and almost nobody writes it down: you can improve two of the three things that produce revenue and still go backwards, because revenue is a product, not a sum.
Disclosure. PlaceForPros is a company I co-own, and AdMax is its marketing agency. Every figure below is pulled directly from the store's Shopify analytics for the stated periods and is published with the company's permission. Nothing here is modelled or estimated.
The comparison window is 1 April to 7 August, 2026 against the identical window in 2025. Same store, same category, same seasonality.
Read the first two lines together, because that is the genuinely good news buried in a bad headline. Traffic nearly halved and orders went up anyway. Whatever was broken between arriving on the site and completing a purchase is materially less broken than it was.
Broken out by channel, the period is not a story about decline. It is a story about one channel replacing another, badly enough on volume and well enough on quality that the two nearly cancelled.
Search went from 7% of all traffic to 33%, and from 15% of revenue to 30%. Its own conversion rate barely moved, 0.68% to 0.64%, which is the number that matters most here: the new traffic converts at roughly the same rate as the old traffic did. It is not padding. That is what a working AEO and GEOprogram looks like in a P&L rather than in a rank tracker.
Meanwhile direct traffic fell 57% and social fell 65%. The store did not lose its funnel. It lost its old top of funnel and grew a new one that was not yet big enough to cover.
Any time you see a report where one channel doubles while another collapses, the honest first question is whether anything actually happened or whether the analytics simply started labelling the same visitors differently. Improved UTM hygiene alone will move thousands of sessions out of "direct" and into "search" without a single new customer arriving.
So here is the test, and it is a good one to keep. If the search gain were reclassification, direct orders would have fallen by roughly the amount search orders rose, because the same purchases would just be wearing a different label.
They did not. Direct orders went 182 → 178, flat. Search orders went 43 → 98, up 55. Total orders went 255 → 287, up 32, with social giving back 19.
Direct held while search nearly tripled its order count. The growth is incremental. If you are going to publish a channel win, run this check first, and publish the check too.
Revenue is traffic multiplied by conversion rate multiplied by average order value. Three motors, and because they multiply, one of them collapsing can swallow two of them improving.
Traffic fell 46%. Conversion doubled, which more than covers it. Together those two produced the +12.5% in orders. Then average order value fell 16.5%, from $377.73 to $315.29, and the +12.5% in orders turned into -6.1% in revenue.
The basket, not the funnel, is the whole story. And it is worth noticing how invisible this is on a normal dashboard. Conversion rate was up and to the right. Search was up and to the right. Orders were up and to the right. Three green charts, and the business made less money.
Hold everything else exactly as it happened and put average order value back to last year's $377.73. The same 287 orders would have produced $108,408 instead of $90,488.
That is the entire prioritization argument in one line. The next dollar of effort does not belong in acquisition and it does not belong in conversion, because both are already working. It belongs in product mix, bundling, and merchandising the higher-ticket end of the catalogue back into the buying path.
It is also, bluntly, the cheaper fix. Doubling a conversion rate takes a site rebuild. Recovering a ticket takes a merchandising decision.
I would rather publish this than leave it out, because leaving it out is how case studies become advertising.
Organic social delivered 8,688 sessions over the four months. Those sessions produced 11 orders and $842, at an average order value of $77. Facebook contributed 3,886 sessions and YouTube 3,407. For a business selling professional tools with a historical ticket near $390, that channel is functionally a zero.
There are two honest readings. Either the mechanic is wrong, meaning the content is reaching people who are not contractors, or the channel is genuinely upper-funnel here and should be measured on assisted rather than last-click. We are testing the first. But four months and $842 is enough evidence to stop defending it on faith, and the effort is better spent on the channel that is demonstrably paying.
The practical takeaway is a habit, not a tactic. Every time you review performance, decompose revenue into its three motors and look at all three, because your dashboard is organised by channel and channels do not tell you which motor is failing.
In this account, the diagnosis inverted the roadmap. Going in, the assumed problem was traffic, because traffic was down 46% and that is the number that screams. Traffic was the least important thing on the list. Conversion was already fixed. The ticket was quietly costing more than either.
A search program that grows revenue 86% while total traffic halves is doing its job. A funnel that doubles conversion is doing its job. Neither of them can save you from a shrinking basket, and no amount of additional acquisition spend would have found that out.
Want AdMax to do the AEO and GEO work for your brand? Book a thirty-minute call with a senior strategist.