Selected engagements
The receipts.
Our engagements are under NDA, so clients are described rather than named. Everything else is published: the sector, the duration, what the problem actually was, and how each number was measured.
Where a figure is still awaiting client sign-off, its basis line says so rather than being quietly dropped.
E-commerce
The problem
Paid social was the only channel producing volume, and it had stopped scaling. Blended CPA had risen for five consecutive months while spend stayed flat, and nobody could say whether the problem was creative fatigue, audience saturation, or a tracking gap after the iOS changes.
What we did
We rebuilt measurement before touching the campaigns: server-side events, a deduplicated purchase signal, and a four-week clean baseline at existing spend. That baseline showed roughly a fifth of conversions were going unattributed, which had been read internally as declining performance. With real numbers in place we restructured around three tested creative concepts and moved budget toward the cohorts that survived a spend increase.
The client’s instinct was that the creative had stopped working. The measurement rebuild showed the creative was fine and the tracking was not. Fixing the second made the first look like it had improved by a fifth before we changed a single ad.
SaaS
The problem
Lead volume was healthy and sales was unhappy. Marketing reported growth every month while the sales team reported that most of what arrived was unqualified. The two functions were measuring different things and each had data to prove it.
What we did
We started by getting both teams to agree, in writing, on what a qualified lead was, then rebuilt the event schema and CRM handoff around that single definition. Reporting moved from form fills to opportunities created. Several channels that had looked strong on the old metric were cut, which reduced raw lead volume in the first quarter before qualified volume climbed.
The first deliverable was not a campaign. It was one sentence defining a qualified lead, signed off by both the CMO and the VP of Sales. Every number after that meant the same thing to both of them.
Local retail
The problem
Fourteen stores, eleven inconsistent Google Business Profiles, and a website with a single contact page listing every branch. Customers searching for the nearest store were finding competitors, and the client had no way to attribute walk-ins to search at all.
What we did
Citation cleanup and profile consolidation first, then individual location pages with genuinely local content rather than a template with the city name swapped. Call tracking numbers per location gave the client, for the first time, a measurable link between search visibility and footfall.
Most of the gain came from unglamorous work: making eleven business listings agree with each other about the client’s own address and opening hours.
Next step
Bring us a number that is not moving.
A strategy session is 45 minutes, costs nothing, and ends with a written view of what we would do first. If that turns out to be something other than hiring us, we will say so.