The AI That Cut Out the Middleman
1st Central used AI to cut its reliance on price-comparison platforms. Alert Electrical used AI to sort its own search data instead of buying more of it. Both cases run against the assumption that AI marketing means less first-party control.
Key Takeaways
- 1st Central Insurance used AI specifically to reduce its reliance on third-party price-comparison platforms, building targeting from its own first-party customer data rather than handing more of the relationship to an outside channel.
- Alert Electrical Wholesalers used a GPT API integration to sort 120,000 monthly search terms into categories the business itself defined, filtering out wasted spend rather than automating a decision away from the company.
- Direct sales rose 33%, and direct quotes rose 50% for 1st Central, while Alert Electrical’s revenue rose 20.77% over three months with 400 more keywords reaching page-one positions.
- In both case studies, the AI is doing the sorting and modelling work, but the categories, the audiences and the campaign structure it works from were built from data the company already owned.
The assumption behind a lot of scepticism about “AI-powered” marketing is that adding AI to a campaign means handing more control to a platform, a black box making decisions a marketing team can no longer see inside or fully own.
Two case studies published on Foregrounded show the opposite. 1st Central, a UK insurance company, used AI specifically to pull business away from third-party platforms and back onto data it already held.
Alert Electrical Wholesalers used AI to sort its own search data into categories it defined, not categories a platform imposed on it. Neither company describes AI as something that took the decision-making away from them.
An insurer trying to stop paying a middleman
1st Central, a car insurance provider, relied heavily on price-comparison channels for new business, carrying acquisition costs that made growth expensive, exactly the dependency a comparison site is designed to create and monetise.
Price comparison websites like Confused.com and GoCompare pair massive marketing budgets with powerful web applications to drive traffic to their websites, offering a simple way for UK consumers to compare prices on Car Insurance, Home Insurance, Loans, and everything in between.
Like many insurers, 1st Central has realised the importance of first-party data and owning the full customer relationship.
Found, its agency, built a proprietary framework called Everysearch, described as an AI-powered approach to the modern consumer journey, and used it to unify PPC brand and non-brand search, Google Performance Max campaigns and demand generation under a single strategy with an explicit goal: reducing 1st Central’s reliance on third-party platforms.
The targeting itself was built from segmented first-party customer data and lookalike audience modelling based on 1st Central’s own high-value customers, alongside Google’s AI-driven audience tools. “Working with Found has significantly improved our customer segmentation and digital marketing at 1st Central. Their AI-powered approach helped us uncover valuable insights from our first-party data, allowing us to deliver highly personalised experiences in paid search and social.
As a result, we have seen impressive results in overall performance,” said David Judic, 1st Central’s Digital and Marketing Director. Direct sales rose 33%, direct quotes rose 50%, and the campaign won a UK Search Award for Best Use of Search, B2C.
A wholesaler filtering its own search data instead of buying more of it
Alert Electrical Wholesalers had an established online presence and more than 10,000 SKUs, but had hit a revenue plateau and switched agencies after outgrowing the support its previous PPC consultants were providing. Imaginaire, the new agency, built a custom integration between Google Sheets and the GPT API to sort roughly 120,000 monthly search terms into five categories the business itself needed distinguished: product, product category, transactional, informational and irrelevant, filtering out spend on queries that were never going to convert.
Using that categorisation, Imaginaire split Alert Electrical’s single Google Shopping campaign into smaller, category-based campaigns, alongside SEO work across the full SKU range.
“We’re delighted with Imaginaire. We chose to switch over because we’d outgrown our previous agency and, thankfully, that’s turned out to be a great decision. Imaginaire offer a great level of customer care and give us great ideas, but most importantly, they drive the result and work to a plan,” said Nathan McEvoy-Swann, Alert Electrical’s Website Manager.
Over three months, revenue rose 20.77%, the site’s keyword footprint grew to 3,000 terms, with 110 more reaching positions one to three and 400 more reaching positions one to ten.
The same underlying move, in two different markets
Read side by side, the two case studies describe the same structural decision from different starting points.
1st Central’s problem was strategic: too much of its new business flowed through a channel it didn’t control, and the AI was pointed at building an alternative, first-party route around that dependency.
Alert Electrical’s problem was operational: too much of an existing, well-established PPC account’s spend was going to search terms that were never going to buy anything, and the AI was pointed at sorting the company’s own data cleanly enough to stop that waste.
In neither case did the AI replace a marketing decision the business used to make itself. In both, it processed a volume of data, 120,000 monthly search terms in Alert Electrical’s case, segmented first-party and lookalike audiences in 1st Central’s, that would have been impractical to sort by hand, and handed the result back to a team still setting the categories, the campaign structure and the strategy around it.
What this means for a marketing team wary of “AI-powered”
Both case studies undercut the assumption that AI in marketing is a step away from first-party ownership. It’s the opposite move in both: 1st Central used AI explicitly to reduce dependence on a third-party platform, and Alert Electrical used AI to make better use of a search account it already owned rather than to introduce a new external dependency.
Neither result depended on 1st Central or Alert Electrical ceding strategic control to Found or Imaginaire’s AI tooling, the frameworks did the sorting and modelling; the businesses kept the judgement calls about what those categories and audiences should be.
For a marketing team treating “AI-powered” as a synonym for “less visibility into how our own campaigns work,” these two case studies are a fairly direct counter-example: in both, AI was the tool used to see more of the company’s own data clearly, not less.
