National considered-purchase retailer.
56,000+ tracked conversions for a single client, with demand capture held through a long research cycle and sustained over years.
single client
A category shopped rarely and researched heavily. Demand capture was built around high-intent search, with remarketing sequences matched to research stage. It ran continuously for years and became one of the highest-volume conversion engines in the portfolio.
The situation
In some categories a brand can afford to lose a buyer and win the next one. This category is different. Customers shop it rarely and research it heavily, and once they have purchased they leave the market for a long time. Every in-market buyer has to be found, then held through a long research cycle up to the moment of decision. A buyer who drifts to a competitor mid-research does not come back.
For a national retailer this gets harder. Demand has to be captured in every market the brand trades in, including the weaker ones, and it has to be captured every week. New buyers start researching all the time.
What was done
The engine was built on two disciplines. The first was high-intent search: full coverage of the queries buyers use when they are genuinely in-market, running permanently and through quiet periods as well as promotions. The second was remarketing sequenced to research stage. Early researchers saw guidance, comparers saw the reasons to choose, and buyers close to a decision saw the case for acting now. The message matched where the buyer actually was. Search captured the demand that already existed, and the sequences kept the brand present while that demand matured.
The program also avoided bursts. Considered categories punish stop-start media, because a buyer who begins researching during a quiet month is lost long before the next campaign launches. The engine ran always-on and was tuned continuously for years. Over that time it learned which queries and which messages moved buyers forward.
Why the number is real
The 56,000+ figure counts tracked conversions: completed actions recorded in the client’s own measurement, accumulated for a single client across the life of the engagement. It excludes clicks, impressions and platform-modelled estimates, and it is not a total pooled across a portfolio of accounts. Every conversion comes from one client and one tracking configuration.
Volume at that scale comes from an engine that stays on and keeps learning across every stage of a long research cycle. Buyers now put many of their research questions to AI assistants as well as search boxes. Staying present through that cycle means being the brand the AI-generated answer cites.
The category is researched over months. Staying present across that whole cycle is what produced the volume.
Buyer research is moving from search boxes to AI assistants. Staying present through that cycle is the work of the AI search practice.
Client name is withheld. The engagement and figures are real, and details are available in conversation.
Two more engagements.
$3M+ tracked revenue
Premium eCommerce retailer
A considered-purchase home category with a five-figure average order value. The work was a full-funnel rebuild across Meta, Google and email: creative testing at volume, tracking rebuilt server-side, and spend reallocated monthly against tracked revenue.
Buyers at this price point wanted proof of quality and provenance before committing. The creative was rebuilt around that.
8x organic growth YoY
Specialist online retailer
Inherited a paid account optimising to the wrong conversion event, with no offline feedback loop. Measurement was rebuilt first, before any spend changed. Paid was then restructured and an SEO program added, and organic growth reached eight times its starting point inside a year.
The account was optimising to the wrong conversion event. Once the tracking fed it real sales, the same budget produced different results.
Let’s talk about what’s next.
For executive advisory, fractional CMO, AI search strategy or speaking enquiries.
[email protected]