The question I get asked sounds almost the same every time. Can we not simply run some ads. The honest answer to that is: technically right away, an account is set up in half an hour.
Except I look somewhere else first. Not at the ad, not at the keywords, not at the budget. At the page that comes after the click. Because that is where it is decided whether the money works or is merely spent. And that is where Google looks too, more thoroughly than most people assume.
Most People Think Ads Are Just an Auction#
The idea is simple and widely held: whoever bids more gets shown more often. Whoever loses out has to raise the bid.
That is not how it works. Google calculates an ad rank, and the bid is only one component within it. Also in there: the quality of the ad, the quality of the landing page, the context of the search, ad rank thresholds and how competitive the particular auction is. So the page behind the click does not sit at the edge of the system, it is part of it.
Which means the reverse is true as well. A weak landing page cannot be offset indefinitely with higher bids. You can buy yourself additional traffic. That does not make the page any better in business terms, and that is precisely the difference between an auction you win and a month that paid off.
Google Has Its Own Requirements for the Destination, and They Are Public#
Before an ad runs at all it gets reviewed, destination included. Google lists its destination requirements publicly, with its own name for every single case.
The ones that actually come up in an ordinary business are a manageable handful. A destination that does not work or was set up incorrectly. A destination that is not accessible in the region being advertised to. A destination that Google's review bot cannot crawl because a server configuration or a line in the robots.txt shuts it out. A displayed address whose domain does not match where the user actually lands. A page whose content mainly exists to serve ads. Navigation that is unnecessarily difficult or frustrating.
Two more are worth a special mention because they sit in the ad rather than on the page: a phone number Google has not verified, and a phone number that does not connect the caller to the advertised business. Both are policy violations, not merely cosmetic flaws.
What stands out about this list is what is not on it. Not one of these points has anything to do with the bid, and none of them is visible to somebody who only looks inside the ads account.
What happens on a violation is less dramatic than many expect. No invoice arrives and no penalty, the ad is disapproved and does not run. That is visible in the account, where it is listed as not servable. For destination violations Google additionally states that an account suspension does not arrive without a warning and a deadline first.
The tricky part is not the visibility, it is who does the looking. Somebody who sets up an account and then waits for the monthly report sees a campaign delivering very little and goes looking for the fault in the budget. The notice has been sitting in the account the whole time, just not where that person checks.
The Ad Makes a Promise and the Page Has to Keep It#
The case most often underestimated is called destination mismatch. What is meant first is something technical: the domain in the displayed address has to match the address the user lands on, redirects must not push them onto a different domain, and a tracking template has to lead to the same content as the actual address. Google states the intent behind it plainly: ads should accurately reflect where a user is being taken.
The interesting part is the human version of the same rule, which needs no policy at all. Anyone clicking an ad has an expectation in their head, and it comes from exactly the line they clicked. If they then land on a page telling them something more general, that is not a rule violation, but it is the same breach. They clicked on a promise and got an overview.
At that moment the click is already paid for. What happens next is the actual difference between a campaign that adds up and one that merely runs.
Quality Score Is Not What Most People Think It Is#
A lot of half knowledge gets passed around here, including by people who ought to know better professionally. The common version goes: a poor landing page lowers your Quality Score, and a poor Quality Score makes every click more expensive.
That is not supported. Quality Score consists of three parts according to Google, and landing page experience genuinely is one of them. About the number itself, though, Google writes that it is a diagnostic tool and is not used at auction time.
Holding both of those as true at once is the whole point. The number in the account is feedback, not a price tag. Landing page quality, meanwhile, is explicitly part of ad rank, which is where position and cost are decided. So the page counts. Just not through the metric everybody points at.
In practice that means: optimising the number is optimising a thermometer. Improving the page changes the temperature.
What Google Actually Assesses on a Landing Page#
If the page is assessed, the next question is fair enough: assessed on what. Google describes landing page experience in its help pages with surprisingly concrete components. How useful and how relevant the information on the page is. How easily a user finds their way around it. How many links are on the page. And what expectation the visitor arrives with, formed by the ad they clicked.
On top of that come two things Google highlights specifically for mobile: that ease of navigation matters even more there than on a desktop, and that a faster landing page is among the most effective levers for mobile ads. And from the policy corner, the point about original content: a page whose content mainly exists to display ads counts as insufficient.
What is striking is how little of this sounds like marketing. Not one of these points is a trick. It is a description of what a person experiences after clicking something and wanting to know whether they are in the right place. Build the page for that person and you satisfy the assessment along the way.
One Page for Everything Gets More Expensive Under Ads, Not Cheaper#
Plenty of businesses have a homepage carrying every service they offer, and send the paid traffic there. It feels economical, because there is only one page to maintain.
In reality it is the most expensive option. Every searcher arrives with exactly one question and meets a list in which their question is one item out of eight. They have to work out where to go next, and most of them do not. They go back. The click is paid for regardless.
There is also the fact that the effect degrades with every additional topic advertised. A page answering one question cleanly can be strong for that one topic. A page brushing past eight questions is mediocre at every one of them. Anyone serious about ads therefore needs a dedicated destination per advertised thing that deals with exactly that thing. This is not a trick, it is the basic condition.
Without Measurement You Are Buying Clicks, Not Customers#
The last blind spot sits at the end of the chain. If there is no way to measure what became of a click, you are buying visits and afterwards only know how many there were.
It starts off mundane. A contact form that validates nothing accepts mistyped addresses and reports success anyway. The enquiry counts as sent, the reply reaches nobody, and the report shows a click that looks like a win.
Then there is the consent layer, which nobody in Europe gets around. Depending on how consent mode is implemented and how a visitor decides, measurement signals can be limited or missing entirely. Google can partly fill such gaps through modelling. Anyone who has wired the consent signals incorrectly is therefore optimising against a data set that does not fully reflect what visitors actually did. That is not a reason to skip advertising. It is a reason to set measurement up before the first campaign rather than after it.
Why I Advise Against It More Often Than For It#
Ads are a good tool. They are fast, they are steerable, and within days they produce insights that would otherwise take months. The problem is never the tool, it is the order.
When somebody asks whether we should run ads, and the answer to the question being advertised does not yet exist on the site, I would spend this month's budget building a page that answers it, and advertise on that page next month. It is the less comfortable order, because nothing blinks in the first month. It is also the only one where the money is still there afterwards, in the form of something that keeps working without a budget.
An ad stops working the moment you stop paying. A page that answers a question does not.
