Skip to content
Chandan KumarCkumar Mehta

Is your Google Ads account actually set up properly?

20 questions, the same ones I work through when an account lands on my desk. You get a score, the band you fall in, and the one area to fix first — which is rarely the area people expect.

Free, no sign-up, about 4 minutes. Answers stay in your browser; nothing is sent anywhere unless you ask for the written breakdown.

Measurement

35% of score

Everything downstream is decided by whether the conversion number is real. Fix this first or every later decision is guesswork.

  • Does every primary conversion represent real business value?

    Not page views, not newsletter signups, not time on site. A lead you would follow up, or a sale.

  • Do your Google Ads conversions reconcile with what the business received?

    Take last month. If Ads says 84 and the team remembers 30, one of those numbers is fiction.

  • Is conversion counting set correctly — 'One' for leads, 'Every' for sales?

    Getting this backwards inflates lead counts and teaches bidding to buy repeat form-fillers.

  • Are phone calls tracked, with a minimum call length?

    A click-to-call counted as a conversion regardless of duration counts misclicks as customers.

  • Do you import offline outcomes — which leads actually became customers?

    Without this the account optimises for leads, not revenue. The two diverge more than people expect.

  • Is GA4 linked, with consent mode configured if you serve the EU or UK?

    Also the single most common cause of quietly under-reported conversions.

Campaign structure

28% of score

Structure decides which searches receive your budget. Poor structure quietly funds researchers instead of buyers.

  • Are campaigns organised by buying intent rather than by product category?

    Researching, comparing and ready-to-buy should not share a budget. There are always more researchers.

  • Has someone read the search terms report in the last 30 days?

    Your keyword list is not what you pay for. The search terms report is.

  • Is there a working negative keyword list, maintained rather than set once?

    A negative list built at launch and never revisited is a list of last year's mistakes.

  • Does your brand traffic sit in its own campaign?

    Mixed with generic terms, cheap brand conversions make the whole account look healthier than it is.

  • Is budget distribution roughly proportional to conversion distribution?

    A campaign taking 40% of spend and producing 10% of conversions is a structure problem, not a bidding one.

  • Is geographic targeting set to 'presence' rather than 'presence or interest'?

    The default setting shows ads to people merely interested in your area. It is rarely what anyone wants.

Landing pages

22% of score

The click is the cheap part. Most of the outcome is decided after the visitor arrives.

  • Does paid traffic land on a dedicated page, not the homepage?

    Sending an ad about one thing to a page about twelve things is the most expensive habit there is.

  • Does the page repeat the visitor's search language above the fold?

    Their phrase, not your product name. They need to know within a second that they arrived.

  • Does the form ask only for what you need for the first conversation?

    Every field costs completions. You can qualify on the call.

  • Is there specific, checkable proof on the page?

    One real number, named certification or photograph of the actual thing beats a row of logos.

  • Does the page load in under about two and a half seconds on mobile?

    Paid traffic is impatient by construction — they did not seek you out.

Bidding & budget

15% of score

Last, deliberately. Automated bidding is good at hitting whatever it is pointed at — pointing it correctly matters more than the strategy.

  • Is your bid strategy matched to your conversion volume?

    Under roughly 30 conversions a month, Maximise Conversions usually beats a tCPA target.

  • Is your target CPA derived from real break-even economics?

    Not from what you wish it were. An unreachable target starves delivery.

  • Do you change one thing at a time and wait two weeks?

    Stacked changes during a learning period make results unreadable.

Want the fix list, in order?

I'll email your score with one message per pillar — measurement, structure, landing pages, bidding — each with the specific checks I run on a client account. Five emails, then I stop.

Your result is already on this page — this only sends the written version. No sharing with anyone, unsubscribe in one click from any email.

Why this order

The four areas are not equally weighted and they are not independent. They form a stack, and each one only works if the one above it does.

Measurement decides whether any number you look at is real. Structure decides which searches receive the money. The landing page decides what happens to the click you paid for. Bidding decides how aggressively to buy more of whatever the first three produced.

Work bottom-up and you optimise bids toward a conversion action that was never real, on traffic that was never right, landing on a page that was never built for it. That is why accounts get rebuilt three times in their first month by people reacting to noise — and each rebuild restarts the learning period.

What to do with a low score

Nothing dramatic. Take the weakest area, take the questions inside it you answered “no” to, and work through them one at a time. Most are settings changes or an hour reading a report. Very little of this work is difficult; almost all of it is the kind that falls off the list when the week fills up.

Now cost it outHave me run it properly