What a Google Ads account should look like at ninety days

Fibvo editorial · 21 August 2026 · 4 min read

The monthly report says conversions are up. The sales team says the enquiries have not improved since March. Ninety days into a Google Ads account, that gap is what matters, and by then there is enough search data to argue with. It usually sits in four places: a conversion action that counts anyone reaching a thank-you page, a campaign structure copied from a template, a negative keyword list last touched at launch, and a report that leads with impressions. Here is what it should look like instead, and how to check each part in an afternoon.

Conversion tracking should count qualified enquiries

This decides everything else, because automated bidding does exactly what you tell it to. Set the conversion action to "form submitted" and the system finds the cheapest people who submit forms. Those are rarely the people who buy.

At ninety days the account should bid on a qualified stage, not a page load. Tag every enquiry with its source and campaign in the system your sales team already uses. Have a person mark it qualified or junk on first read. Send the qualified ones back to Google as an offline conversion with a value attached, and bid on that. A tap on the number is not a conversation, so gate call conversions at sixty or ninety seconds of connected time.

• One primary conversion action per campaign goal. Everything else secondary, so it reports but does not steer bidding.
• Qualified enquiries imported inside the click window Google allows. Late imports are dropped quietly and nobody tells you.
• Deduplication set deliberately: one conversion per enquiry for lead generation, every conversion for retail.
• A monthly reconciliation between Google's conversion count and the enquiries in your CRM. An unexplained gap is a tracking fault, not a rounding difference.

An account that bids on form fills gets better at form fills every week, and no better at customers.

Campaign structure should follow intent and margin

Budget is set at campaign level, and so is most of your control. Anything you want to fund or starve on its own needs its own campaign. Product categories are usually the wrong dividing line. What the searcher wants, and what the sale is worth, are the right ones.

• Brand searches in their own campaign. Cheap brand conversions otherwise flatter everything else and hide what cold traffic costs.
• High-margin work separated from low-margin work. A shared budget quietly funds whichever gets more clicks, not whichever pays better.
• Geography split where delivery cost, licensing or lead time genuinely differs.
• Search kept separate from Display and Performance Max, so you can still tell which produced the enquiry.
• Enough volume in each campaign for automated bidding to learn from. Over-segmentation starves every box you create.

Broad match with a reliable qualified-enquiry signal behaves reasonably. Broad match with a form-fill signal spends the budget on curiosity. Earn the signal first, then widen.

Negative keywords should be a weekly habit with a paper trail

The search terms report is the only place the account tells you what it bought. Ninety days in, it should show steady pruning rather than one burst in week one.

• A weekly pass through search terms, filtered to spend with no qualified enquiry.
• Shared negative lists for the permanent categories: jobs and careers, free, DIY, training courses, wholesale, countries you do not serve.
• Campaign-level negatives to keep traffic in the right campaign; ad-group negatives only to steer between close variants.
• Phrase and exact negatives in preference to broad ones, each checked against your own converting terms first. One careless broad negative can silence a campaign that was working.
• A dated log of what was added and why, so nobody reinstates a term you blocked on purpose.

Landing pages should answer the search, not the category

You have already paid for the click by the time the page loads. Sending every ad group to the homepage, or to a category page listing nine services, wastes the most expensive traffic you buy.

• The headline repeats the ad group's keywords in the words the searcher typed, not your product names.
• One primary action per page, visible on a phone without scrolling.
• A form asking only what your team needs to make the first call. Every extra field is a filter you did not intend to install.
• A price range wherever you can honestly give one. Pages that hide it collect enquiries from people who leave when they hear the number.
• Load time measured on a mid-range Android phone on mobile data, not on the office connection.
• A thank-you page that fires the conversion once and tells the enquirer when a reply will arrive. We reply to every enquiry within one working day, and our pages say so.

The reporting line that matters

One line carries the account: cost per qualified enquiry, trended month on month, next to the count of qualified enquiries and the target written down before work started. Once the sales cycle allows it, add closed value against spend. Impressions, clicks, click-through rate and quality score are diagnostics. They explain a movement in that line; they never lead the report.

Two structures keep it honest. The ad account, the conversion data and the analytics property sit in your own name and billing, so the history moves with you if you change agency. Whoever manages the account takes no commission from any platform they recommend, so advice on where to spend stays about your margin rather than theirs. FIBVO works this way, and we report the months that miss target as plainly as the months that beat it.

If the account fails most of this at ninety days, the spend is not wasted. The search terms report alone is worth more than the clicks were. Fix the conversion signal, restructure around it, and treat the next ninety days as the first that count.

---

Written by Fibvo — Bengaluru and Erode, India — working with clients in other time zones. Questions, or think we have it wrong? Tell us.