What you need before you start Google Ads
by Harry Watts · · updated · 6 min read
Google will happily take you from sign-up to live campaign in twenty minutes. That speed is a feature for Google and a trap for you, because most "Google Ads didn't work for us" stories were decided before the first pound left the account, by what wasn't in place.
This isn't gatekeeping; it's sequencing. Four preconditions, in the order they compound, and then, because the industry loves inventing prerequisites, the list of things you genuinely don't need.
1. An offer a stranger would choose
The precondition nobody writes about, because there's no setting for it. Ads put you in front of someone comparing three tabs. If you can't say in one sentence why they should pick you over the established competitor whose ad sits above yours (clearer price, faster response, narrower specialism, better guarantee), the auction will charge you full price to come third.
The quick test: does anything convert strangers now? Word-of-mouth converting isn't evidence: those people arrived pre-sold. If cold visitors from any source never become customers, ads amplify silence, expensively. Sharpen the offer first; it's free, and it improves everything downstream.
2. A page that can convert full-price traffic
Every click is bought at market rate whether the page does its job or not. Ready means four things, none of them design awards:
- Fast on a phone: a couple of seconds, tested on mobile data, not office wifi.
- Message match: the page says what the ad promised, in the first screen.
- One obvious next step: a short form or a visible phone number. Not five menu options.
- Reasons to trust: real address, reviews, specifics. Strangers audit harder than referrals.
Homepage or dedicated page? Honest answer: a good homepage beats a bad landing page, and a dedicated page matched to the campaign beats both. Start with whichever of those you can have ready this month.
3. Measurement wired, then tested
Before launch, not after: retrofitting tracking means the learning period runs blind and the early spend teaches nobody anything. Three steps, one afternoon:
- Conversion actions for real outcomes only: enquiries, sales, calls with duration. No page views, no button clicks.
- A test enquiry through your own site, confirmed as recording exactly once. If it doesn't land cleanly, that's the launch task.
- Consent signalling that works: a banner that blocks tags without Consent Mode quietly deletes the very data you're about to pay to generate.
4. A budget derived, not guessed
"What should I spend?" has an arithmetic answer: customer value → affordable acquisition cost → clicks per customer → affordable CPC, compared against your market's actual click prices. If the sums don't clear, no budget is the right budget yet.
If they do clear, add the comfort test: picture the derived budget spent over a fortnight with zero sales, which is a normal start, not a failure. If that picture would panic you into switching everything off, start smaller or wait a quarter. Panic-editing in week two does more damage to new accounts than any setting Google ships, because it resets the learning you just paid for.
The launch-day settings, so they're in one place
- Auto-apply recommendations off
- Location options set to presence only; Display Network unticked on Search campaigns
- The weekly search-terms half hour already in the diary
- Your success number (target cost per customer), written down where you'll see it, dated, with the 60–90 day judgement window next to it
What you don't need
The industry has a financial interest in a long shopping list. For the record, you can launch well without:
- A rebrand, new logo, or "brand guidelines": strangers judge the offer and the page, not the kerning
- A blog, social media presence, or video assets: nice eventually, irrelevant to a search campaign
- A large keyword list: a handful of high-intent terms beats five hundred maybes
- Performance Max: earn your way into automation with conversion data first
- An agency retainer: a proper build plus your own weekly attention is a legitimate way to start
Complexity is something a working account earns its way into. It's not an entry requirement.
Quick answers
What needs to be in place first?
An offer a stranger would choose, a page that converts, tracking installed and tested, and a budget derived from arithmetic. In that order: each one compounds the next.
Does the website really need to be ready first?
Yes. Ads charge full price for every click the page wastes — a weak page doesn't make ads underperform, it makes them measure the leak precisely.
Tracking before the first campaign?
Always. Real outcomes only, a test enquiry recorded exactly once, and consent signalling that isn't silently deleting data. If the test doesn't land, that's the launch task.
How do I set the first budget?
Derive it from customer value and your market's click prices, then apply the comfort test: if a fortnight at that spend with zero sales would panic you, go smaller. Panic-editing beats every other way of wrecking a new account.
What don't I need?
A rebrand, a blog, video, a huge keyword list, Performance Max, or a retainer. One tight campaign on a converting page with honest measurement beats an elaborate launch.
Want the four preconditions checked, or built
The account setup service is exactly this list done properly: offer sharpened in the ad copy, tracking built and tested, defaults corrected, then handed over with the weekly routine. Or start with a free intro call: thirty minutes against these four headings tells you whether you're ready to spend, and if you're not, which one to fix first, which costs nothing to hear.