What Google Ads really costs when your site is slow.

Fix the page first. Running Google Ads to a slow landing page charges you twice: once for the clicks that leave before anything renders, and again on every remaining click, because the auction prices your ad partly on how the page behaves. It is the one mistake in paid search where a bigger budget makes the loss bigger.

That is not an argument against advertising. It is an argument about sequence. The page is the cheaper half of the job and it has to come first, and this page shows the arithmetic rather than asking you to take it on faith.

You are buying sessions, not clicks

A click is a charge. A session is a person who got far enough to read something. The gap between the two is pure loss, and it is the only line in a paid search account that nobody reports on.

Picture the moment properly. Somebody is standing in a driveway on a four-year-old phone, two bars of signal, mild panic about whatever has broken. They tap your ad. Nothing happens for a few seconds, because your hero image is three megabytes and a chat widget is loading from somebody else's server. They go back to the results and tap the next advertiser.

Google charged you for that. The next advertiser got the enquiry. Nothing in your dashboard calls this out, because the click was delivered exactly as promised.

That is why the first question we ask about an ads budget is not what the budget is. It is what the landing page does on a phone on mobile data.

The arithmetic on a $40 click

Competitive trade and professional service terms genuinely clear $40 a click in Ontario, so use that as the working number and swap in your own once you have it.

Assume $4,000 a month of spend at $40 a click. That buys 100 clicks. The only variable that changes below is how many of those 100 people stay long enough to be reachable.

Those two abandonment shares are assumptions made to show the shape of the problem, not measurements of your site. Put your own in. The point survives any reasonable numbers you choose: the slow page pays a third more per enquiry for identical traffic, and a thousand dollars a month disappears before anybody reads a word.

Now notice what happens if you respond to the right-hand column by raising the budget. At $8,000 a month the waste doubles to $2,800. The campaign looks like it is underperforming, so the obvious lever is more spend, and more spend is the one lever that makes this specific problem worse.

Two other costs that do not fit in the table. Every form you never receive is a customer who hired a competitor and will not be back next year. And the data you are optimizing on is distorted, because Google's bidding learns from conversions, and a page that loses a third of its visitors teaches the algorithm that your good keywords are bad ones.

What Quality Score actually does to your cost per click

Here is the part that surprises owners, because it means the slow page raises the price of the clicks that do survive.

Your position in the auction is not decided by your bid alone. Google works out an ad rank from your bid together with quality signals and the context of the search. Because quality is a multiplier rather than a tiebreaker, a strong advertiser can buy the same position for less, and a weak one has to bid up to stand in the same place. You can be outbidding a competitor and sitting beneath them.

In the interface this surfaces as Quality Score, reported from 1 to 10 at keyword level, with three components each rated below average, average or above average:

• Expected click-through rate. Whether people tend to tap this ad for this search.
• Ad relevance. How closely the ad text matches the query.
• Landing page experience. What happens after the tap.

Treat the number as a diagnostic readout rather than the thing itself. Google is clear that the live auction uses its own real-time quality signals, and that the reported score is there to tell you where to look. What matters for your budget is the direction: when landing page experience sits at below average across your main keywords, you are paying a handicap on every click in the account, all month, silently.

The useful habit is to add the three component columns to your keyword view and leave them there. Most accounts we are shown have never had them turned on, which means nobody has looked at the one report that explains the cost per click.

Landing page experience, in plain terms

Speed is part of this and it is not all of it. Google's own guidance on landing page experience is about the whole arrival, and it maps closely onto what a customer would complain about.

• Does the page answer the exact search? An ad for furnace repair that lands on a general home page is a mismatch, no matter how fast the home page is. One ad group, one promise, one matching page.
• Is it usable on a phone? Tap targets you can hit with a thumb, text you can read without pinching, a form that does not demand a postal code before a phone number.
• Is it honest about who you are? A real business name, a real address or service area, a phone number, terms. Thin pages with none of that read as risky to a visitor and to a reviewer.
• Can people get around? A click should not land in a dead end with no way to see proof, pricing or anything else.
• Does it load before patience runs out? The specific measures, the thresholds and the four usual causes are in Core Web Vitals for a small business site.

None of that is exotic. It is the same list that makes a page convert when the traffic is free, which is why the fix is never wasted even if you stop advertising.

What to fix before you raise the budget

In order of return for the effort. The first three are usually an afternoon.

1. Compress the hero image on every page an ad points at, and make sure it is not lazy loaded.
2. Delete the third-party scripts you are not using. Old pixels, an unanswered chat widget, two analytics tools measuring the same thing.
3. Put the phone number in the thumb zone as a tappable link, at the top, on mobile.
4. Cut the form to the fields you genuinely need to quote the job, and nothing else.
5. Match the page to the ad group. If you advertise four services, you need four pages, not one.
6. Add the proof that closes a stranger. Photos of your own work, reviews, service area, licensing.
7. Confirm conversions are firing correctly before you judge any of it, because half the accounts we audit are optimizing on a conversion that counts page views.

Then raise the budget. In that order the extra spend lands on a page that converts, and the quality signals improve while you sleep.

When we would run the ads anyway

Being honest about the exceptions is part of the argument.

Run them when the campaign is a phone call. If the ad uses call extensions or a call-only format and the customer never reaches the site, the page matters far less, and for emergency trades this is frequently the best structure regardless of page quality.

Run them when the season is now. A roofer in the middle of a storm week, a tax preparer in March, an HVAC business in a heatwave. A perfect page in November is worth less than a mediocre page during the fortnight your customers are actually buying. Advertise, bank the work, rebuild in the quiet month.

Run them when the page is merely plain rather than broken. Plain and fast converts. Pretty and slow does not. If the page is honest, quick and single-minded, it does not need to be beautiful before you spend.

And do not run them when you have no way of knowing whether the phone rang because of the ad. That is a tracking problem rather than a page problem, and it costs the same amount of money.

The order of operations

• Measure the page, on a phone, on mobile data, at the URL the ad actually points at.
• Fix speed, the form and the match between ad and page. Most of this is hours, not weeks.
• Confirm conversion tracking counts something a customer did, not something a browser did.
• Start the spend small, at the budget where the data is still readable.
• Watch landing page experience alongside cost per enquiry, not clicks.
• Scale once cost per enquiry is stable, not once the click volume looks good.

If you want to see where the paid side sits in the wider picture, that is what we do on the paid media side, and the honest version of the sequencing argument also appears in what a website costs in Brantford, where fixing the site comes before buying traffic for the same reason.

---

We would rather quote you the page fix than the ad management, in the cases where the page is the thing costing you money. Send us the campaign and the landing page and you will get the arithmetic above with your own numbers in it.