First month in a new Google Ads account, day by day

A client wanted to start with €1,500 per month. I opened the website and the first thing I saw was an 11-field form, on a page that took eight seconds to load on a phone.

We started two weeks later. The discussion was unpleasant and honest.

Below is what the first month usually looks like, split by week, for a small service business or a small shop.

Before day one

Conversions, installed and manually verified: form, phone click, WhatsApp click, each as a separate event. On the phone, calls are often the main channel; if you don't measure them, the campaign looks weaker than it is.

Cookie consent, genuinely functional, not decorative.

Landing page, checked on a real phone, on mobile data. Not in the browser on a computer.

Negative keyword list, started before the first click: "free", "job", "employment", "course", "pdf", "second hand" — this changes depending on the domain.

Week 1: start small and check often

One search campaign, two or three groups divided by intent: the service, the service plus city, the price. Ads written separately for each intent.

Just one city, if the budget is small. I once started in five cities at once and got thin data from everywhere, from which nothing could be decided.

Bidding on clicks or manual at first. Target cost per conversion only when we have consistent conversions, not a peak on one day.

We look at the search terms report twice this week. That's where you see what the money is being spent on. For one client, I found clicks on searches related to a recruitment competition in another county.

Week 2: we clean up

We add negative keywords from what we found. We stop what doesn't make sense. We check if the incoming enquiries resemble the clients the client wants; if not, the problem is in the keywords or the ad, not the budget.

We also check the boring part: do notifications reach the client, does someone call about the enquiries, and how quickly. A form that sends to an unread inbox is worse than no form at all.

Week 3: we test ads, not settings

Headlines that state the result and the area. An ad that answers the most frequent objection. An ad that talks about price, even if the answer is "it depends, we'll tell you on the phone in 10 minutes".

Extensions completed: location, call, links to service pages. Not for score, but because they take up space in the result and offer options.

Week 4: first budget decision

We look at cost per enquiry, per channel, and what happened after the click. If traffic comes in and produces nothing, increasing the budget only magnifies the problem.

We increase the budget in steps, not double overnight. When we suddenly doubled, the system re-entered learning and the cost went up for a few days. We got burned once like that and since then we increase slowly.

We postpone remarketing if the budget is small. It would eat up money that does a better job in search.

What we stop and when

A group that spends money without any enquiry for several weeks, with enough click volume not to be just bad luck. A keyword that consistently brings unsuitable enquiries. Sometimes, the whole account — if the site isn't ready.

We don't have a fixed stop figure; it depends on the niche and how much an enquiry costs for you. We have a rule to write down at the beginning how much we are willing to lose on a test, so we don't change our minds under pressure mid-way.

What we report

Enquiries and cost per enquiry, compared to what the client knows from their accounting, not just from the platform. What new terms appeared. What we stopped and why.

What we don't report as an achievement: impressions, average position, click percentage without context.

If you already have an account and want an opinion, give us read-only access. We'll send you a list of what we would clean up, without obliging you to work with us.

How we split the budget between campaign types

We don't have a recipe with percentages, because it depends on how well-known the company is and what it sells. But we do have an order in which we start.

We intentionally leave automated campaigns for the end. Without clean data, they learn incorrectly and it's hard to fix afterwards.

The monthly report, what it contains for us

Incoming enquiries, cost per enquiry, what we changed, and what's next. We don't put impressions and clicks on the first page, because they don't decide anything for the client. When a month is weak, we write why we think it was weak, even if the reason is a decision of ours.

A report where everything is green, month after month, is a report nobody reads carefully.

What we ask from the client along the way

To tell us which enquiries were serious and which weren't. Without their feedback, we optimise based on a number, not on quality. For one client, we cut a group of terms that brought many enquiries, because none turned into work — and only they could tell us that.

What we don't promise in the first month

We don't promise a fixed cost per enquiry and we don't guarantee volume. The first month is, for the most part, paid for us to find out what works. We say this before starting, so it's not a surprise in the first report.

If you want us to start a new account without burning through the budget in the first week, let's have a quick call and we'll tell you what we'd start with.