The goal was simple to state and hard to hit: turn a coaching institute's monthly Google Ads spend into actual admissions, not just clicks. They were already spending a fair bit every month. The money was going out; the quality leads weren't coming back. Cost per lead kept drifting up, hardly anyone enrolled, and nobody on their side could say why.
So before touching a single bid, we asked for four weeks of evidence: search query reports, conversion attribution, call recordings from a sample of leads, and the rejection reasons for inquiries that never became admissions. Reading through it, three things jumped out. A chunk of the clicks were job seekers typing coaching-related queries looking for employment, not students or parents. One landing page was doing the work of four different exams and converting poorly at all of them. And the inquiry form asked for so little that the sales team walked into every call blind.
Structuring the account around exams
We tore down the flat structure and rebuilt it by exam (UPSC, SSC, Bank PO, State PSC), each with its own ad group, keyword set, ad copy and landing page. Four dedicated pages went up, one per exam, each carrying testimonials, the faculty who actually teach that exam, and the real class timings. A UPSC aspirant and a Bank PO aspirant want different things; now the messaging matched.
Filtering out the tyre-kickers
This is where most of the waste was hiding. We built a thorough negative keyword list to block job-related queries, free-coaching hunters and irrelevant exam categories. Conversion tracking went in properly for both form submissions and phone calls, with call-duration filtering, so a six-second misdial didn't get counted as a lead. Then the inquiry form: expanded to capture target exam, intended start month and budget, giving the sales team context before they dialled instead of after.
On targeting, we leaned into where they were actually strong. Location bid adjustments pushed harder in the Tier-2 cities where the institute had a reputation, and pulled spend back in the metros where bigger national players dominate the auction.
Why we waited four weeks to switch bidding
Only after the negatives were live and the tracking was clean did we move from manual CPC to target CPA. Smart bidding is only as good as the conversion data feeding it. Flip the switch too early and you're handing Google your noise to optimise against. So we sat on manual for four weeks on purpose.
The results held up over a few months. Cost per lead came down, the volume of genuinely qualified leads grew, and because the leads were better, more of them turned into admissions. Spend stayed roughly where it started. The gains were efficiency, not a bigger cheque. Honestly, none of this was clever. Most coaching accounts bleed money on the same three things: broad match with no negatives, one page for many courses, and tracking nobody trusts. Fix those and you beat the "advanced" tactics every time.
