Scripts that gave a specialist team its time back

The situation

Confused.com runs one of the largest paid search operations in UK insurance, with its own highly technical in-house team. Accounts at that scale generate more routine checking than any team can do by hand: errors to catch, settings to keep tidy, search terms to sift, keywords to review. It can be relentless. In 2019 the team asked us to help take that load off with automation assistance.

What we did

We worked alongside the in-house team and wrote customised Google Ads scripts, well before the explosion in language models, that queried the accounts programmatically and did the repetitive work at scale.

• Account hygiene checks, run on a schedule rather than when someone remembered.
• Error checking and alerts, so problems surfaced the day they appeared.
• Keyword review across the accounts, flagging what needed a human decision.
• N-gram analysis of search terms, to find the words and phrases driving cost and conversions that no one would spot term by term.
• Queries that looked for opportunities and pressures across the accounts and reported them, so the team could act rather than hunt.

How we measured it

By counting what the team no longer had to do daily by hand. Each script replaced a recurring manual check, and the time that freed went back to the in-house specialists.

The outcome

Their routine checks ran deterministically on code, at scale and on schedule, so the specialists' time went on the decisions only they could make.