Most business owners we speak to can tell us what they spend on Google Ads. Far fewer can tell us what it returns. That gap is not a reporting problem — it is usually a measurement problem, and it is the single most common reason good budgets get cut and bad budgets get renewed.

How to Tell If Your Google Ads Budget Is Actually Working

Here is the sequence we work through when we audit an account, in the order that actually matters.

1. Decide what a result is before you look at any number

A “conversion” in a Google Ads account can be a form submission, a phone tap, a PDF download, a newsletter sign-up, or someone scrolling past 75% of a page. Accounts routinely count all of these together, then report a cost per conversion of AED 40 and call it a success.

Write down the one action that means a real commercial opportunity for your business — a quote request, a booked call, a completed checkout. Everything else is a secondary conversion. It can still be tracked; it just should not be counted in the number you use to make budget decisions.

2. Separate cost per lead from cost per customer

Cost per lead is the number the ad platform can see. Cost per customer is the number that decides whether you keep spending. The difference between them is your close rate, and it is rarely the same across campaigns.

A branded search campaign might produce leads at a third of the cost of a competitor campaign, and close at three times the rate — the same spend, wildly different economics. Until you can attach an outcome to a lead, you are optimising the cheaper half of the equation.

The practical fix is unglamorous: get a lead identifier (a GCLID, a form ID, a call reference) into your CRM at the point of enquiry, and get the eventual outcome back out of it. Most CRMs do this with a hidden field and a weekly export.

3. Check what your conversion window is hiding

Default attribution windows in Google Ads are generous. A 30-day click window means a conversion recorded today may belong to spend from four weeks ago. If you compare this month’s cost to this month’s conversions in a growing or shrinking account, the ratio will be wrong in a predictable direction — it flatters growth and punishes cuts.

For considered purchases, compare cohorts: spend from a given week against conversions attributed to that week’s clicks, not to that week’s calendar days.

4. Look at search terms, not keywords

The keyword is what you bid on. The search term is what someone actually typed. In broad match campaigns — which is now most campaigns — those two diverge quickly, and Performance Max reports search terms only in aggregate.

Pull the search terms report for the last 90 days and sort by cost. The exercise is not to find one embarrassing term; it is to see what proportion of your spend went to queries that describe a problem you solve versus queries that merely share a word with it. In accounts that have not been reviewed for a year, we typically see 20–40% of spend sitting in the second group.

5. Test whether the ads are the constraint

If your click-through rate is healthy, your search terms are relevant, and your cost per lead is still too high, the ads are probably not the problem. The landing page is.

Two diagnostics separate the two quickly. First, compare your click-through rate to the auction insights benchmark for your category — that tells you whether the ad is earning attention. Second, compare landing page conversion rate to the rest of your site — that tells you whether the page is converting the attention it gets. A campaign with a 6% click-through rate and a 0.8% landing page conversion rate does not need new ad copy.

6. Judge the account on trend, not on last week

Weekly paid search data is noisy. In a business generating fewer than about 50 leads a month, a single good week and a single bad week will differ by more than most optimisations will ever move the number. Look at rolling 4-week or 8-week averages, and set a review cadence you can hold — monthly for most accounts, weekly only where volume genuinely supports it.

What good looks like

You should be able to answer four questions without opening the ad platform:

• How many genuine enquiries did paid search produce last month?
• What did each one cost?
• What share of them became customers?
• What was that worth?

If any of those four takes more than a few minutes to answer, that is the thing to fix first — before budgets, bids, or creative. Everything else in a paid account is easier to argue about once the measurement is settled.

If you would like a second opinion on an existing account, we run a structured review and send back what we would change and why. Request a proposal and tell us a little about the account.