Account and tracking audit
Structure, match types, negatives and whether your conversions mean anything.
Most underperforming ad accounts are not underperforming because of bidding. They are underperforming because the structure buries intent: one campaign carrying twenty unrelated products, broad match with no negative list, a conversion action that counts a page view, and every rupee judged on a number that does not correspond to revenue.
We fix the foundations first: what counts as a conversion, whether it is tracked accurately, and how campaigns are grouped by intent and margin. Then creative and landing pages, then the daily hygiene of search terms, placements and audiences. Reporting is cost per qualified lead and return on spend.
Most underperforming accounts have a definition problem
When an account is spending well and returning badly, the instinct is to look at bids. The cause is almost always further upstream: the account is optimising towards something that is not a sale. A conversion action counting page views, a thank-you page firing on a bounce, a phone click counted without a connected call. Feed any of those to a bidding algorithm and it will faithfully find you more of exactly the wrong traffic.
So the first fortnight of any engagement is spent on measurement. What counts as a conversion, is it tracked accurately, does it deduplicate, and does its value reflect what that action is actually worth. Everything after that is comparatively easy.
Structure follows intent and margin
A single campaign carrying twenty unrelated products spends its budget on whichever ad group gets the most clicks, not on the one with the best margin. Grouping by intent and by margin puts you back in control of where money goes.
That usually means separating branded from non-branded so the numbers stay honest, separating high-intent from research terms so they can carry different targets, and separating your best products from the rest so they are not competing for the same pot.
Automation is a tool with inputs you control
Smart bidding and Performance Max genuinely work when they are fed accurate conversion data, adequate volume, strong creative and clean audience signals. Fed poor inputs, they will confidently spend your budget and the platform's own recommendations will encourage more of it.
We keep auto-apply switched off, review every recommendation on its merits, exclude brand traffic from campaigns that would otherwise cannibalise it, and hold manual control where volume is too low for a bidding algorithm to learn anything.
The daily hygiene nobody sees on a dashboard
Search term reviews and negative additions. Placement exclusions. Audience and device adjustments. Geography checks. Budget pacing so the month does not end with an unspent third or an overspent first week. Disapprovals cleared before they cost a day of impressions.
None of it is clever and all of it compounds. It is the difference between an account that improves quarter on quarter and one that quietly decays between reviews.
Reporting on cost per qualified lead
Cost per lead is easy to improve and easy to fake: loosen a form, count a newsletter signup, and the chart improves while the sales team gets nothing. Cost per qualified lead requires your team to tell us which leads were real.
We set that feedback loop up deliberately: a simple qualification field in your CRM, imported back into the ad platforms where the integration allows, so bidding optimises towards the leads that actually convert and reporting reflects the business rather than the platform.