From Zero to Paid: Running Your First Google Ads Test Without Burning Your Runway

1. Why Google Ads Still Matters (Even Early On)

Let’s start with the question most founders ask first: with AI answering more and more questions, are search ads still worth it? Yes. The way people search is changing, but the reason they search isn’t.

When someone types ‘invoicing software for freelancers’ into Google, they aren’t browsing. They have a problem and they are looking for something to buy. Search ads let you show up in that moment, today, not six months from now when your SEO starts to rank.

What’s changing? The ad slot is moving with the user. Google now shows ads inside AI Overviews and AI Mode, not just above the regular results. So even as people shift to asking longer, more conversational questions, there is still a place for your ad next to the answer.

But the real reason to run ads early isn’t reach. It’s speed of learning. SEO is a slow build: you publish, wait, and learn over months. Ads compress that into weeks. Within a few weeks of spend you know which keywords bring buyers, which messages get clicks, and what a customer actually costs you. Those learnings feed straight back into your SEO, your landing pages, and your positioning.

Common myths

Ads are only for companies with big budgets

You don’t need a big budget. You need enough budget, focused on a small set of high-intent keywords. The mistake is spreading a small budget across too many keywords and campaigns, so nothing gets enough data to tell you anything.

We tried ads and it didn’t work

Most failed first attempts are a setup problem, not a channel problem. No conversion tracking, broad keywords, ads pointing to the homepage, and every Google recommendation accepted. A test that wasn’t run properly isn’t a verdict on the channel.

Ads and SEO compete with each other

They work together. Ads tell you quickly which keywords convert. SEO makes those same keywords cheaper over time, because you stop paying for every click. Running both early is how you find out where to invest.

The Test Mindset

Treat your first 90 days on Google Ads as an experiment, not an always-on cost. Before you spend a single dollar, decide what a good result looks like: a target cost per signup, trial, or demo that your business can afford. At the end of the test there are three outcomes: scale it, fix it, or stop it. All three are good outcomes, as long as they are backed by data.

2. Understanding What to Bid On

Intent still beats volume

If you have read my SEO guide, the intent types will look familiar. The difference is the order. With SEO, you can afford to build top-of-funnel content for the long term. With ads, you pay for every click, so you start at the bottom of the funnel, with the people who are closest to buying.

Search volume is still useful, but for ads the numbers that matter more are cost per click and intent. In Google Keyword Planner, look at the ‘Top of page bid’ range next to each keyword. You will often find that the broad head term, like ‘crm’, costs more per click and converts worse than a specific one, like ‘crm for small sales team’. The specific keyword has less volume, but the person searching it knows exactly what they want.

Match types in plain language

Match types decide how closely a search has to match your keyword for your ad to show.

• Exact match [crm for startups]: Your ad shows for searches with the same meaning. Most control, least reach.
• Phrase match “crm for startups”: Your ad shows for searches that include the meaning of your keyword, like ‘best crm for startups in india’.
• Broad match (crm for startups): Your ad shows for anything Google thinks is related. Most reach, least control.

For a first test, start with exact and phrase match. Broad match relies on Google’s automated bidding and conversion data to find good searches, and a new account doesn’t have that data yet. Once you have conversions coming in, you can test broad match deliberately. (One trade-off to know: Google recommends broad match or AI Max to show in AI Overviews. That’s a later-stage optimisation, not a day-one setting.)

Negative keywords from day one

Negative keywords stop your ad from showing for searches you don’t want to pay for. Add a starter list before you launch. Common ones for software: free, jobs, careers, salary, course, tutorial, login, template, and open source (unless you are open source). You will add more every week from the search terms report, which we cover in section 6.

Competitor keywords

Bidding on a competitor’s name can work well, because the searcher already knows the category and is evaluating. It works when you have a clear reason for someone to switch and a comparison page to back it up. It works poorly when your product looks the same as theirs. Two things to keep in mind: these clicks can be expensive, and you shouldn’t use a competitor’s trademarked name in your ad text.

How to find the right keywords

• Conversations with customers: The words customers use to describe their problem are the words they type into Google.
• Communities (Reddit, Slack, Discord): Watch how people ask for recommendations in your category.
• Google Keyword Planner: Free with a Google Ads account. Use it for volume and bid ranges.
• Google Ads Transparency Center: Search any competitor and see the ads they are running. It’s a quick way to learn their messaging and which landing pages they send traffic to.

3. Setting the Foundation

Before you spend anything, get the foundation right. Most wasted ad budget comes from skipping this part.

Account setup

When you create a Google Ads account, Google will push you towards a simplified ‘Smart’ campaign. Switch to Expert Mode instead; it gives you control over keywords, bids, and settings. Create the account with your own business email so your company owns it. If you work with a consultant or agency later, give them access to your account. Don’t let them run your ads from theirs.

Settings to switch off at the start

• Search Partners: Shows your ads on sites outside Google Search. Harder to judge quality; turn it off for the test.
• Display Network (inside a Search campaign): Shows your ads as banners across websites. Different intent entirely; turn it off.
• Auto-apply recommendations: Lets Google change your campaigns automatically. Turn it off and review recommendations yourself.
• Location targeting: Change it to ‘Presence’ (people in or regularly in your target locations), not ‘Presence or interest’.

Conversion tracking

This is the single most important step. Without conversion tracking, Google can only optimise for clicks, and clicks don’t pay your bills. Here are the steps:

1. Set up Google Analytics 4 and Google Tag Manager on your website, if you haven’t already.
2. Decide what counts as a conversion. Your primary conversion should be the action that matters to your business: a signup, a trial started, or a demo booked. Things like pricing page views can be secondary conversions, tracked but not optimised for.
3. In Google Ads, go to Goals, then Conversions, and create a conversion action for your primary goal.
4. Install it through Google Tag Manager using the Google Ads conversion tag, or import the conversion from GA4.
5. Turn on enhanced conversions. It improves how accurately conversions are measured.
6. Test it. Complete a signup yourself, check that the tag fires using Google Tag Assistant, and make sure the conversion shows as recording in Google Ads before you launch.
7. Link Google Ads with GA4 and Google Search Console so you can see paid and organic data side by side.

Budget maths: work backwards from your target

Most founders pick a budget by asking ‘what can we afford to spend?’ The better question is ‘what do we need to spend to get an answer?’

Start with your target cost per conversion: how much you can pay for a signup, trial, or demo and still make the numbers work. Then remember that Google’s automated bidding works best once a campaign has a steady flow of conversions, often cited as around 30 a month.

Here is an illustrative example. If your target is $100 per demo booked, 30 demos a month means about $3,000 a month in spend, at your target cost. In the first few weeks you won’t be at target yet, because you are still learning, so you need headroom on top. That’s how you end up at roughly $5,000 a month for a meaningful test. If your budget is much smaller than your maths suggests, it’s better to narrow the test to fewer keywords than to run everything thinly.

4. Your First Campaign: What to Build

Keep the structure simple

Start with one Search campaign. Inside it, create three to five ad groups, each with a tight theme based on intent. For example:

• Category: ‘[category] software’, ‘[category] tool for [audience]’
• Alternatives: ‘[competitor] alternatives’, ‘[competitor] vs’
• Use case or persona: ‘[category] for [industry or role]’
• Brand: your product name

Keep each ad group to a small set of closely related keywords. The tighter the theme, the more relevant your ad, and the more relevant your ad, the less you usually pay per click.

Writing the ads

Google’s main ad format for search is the responsive search ad. You give it up to 15 headlines (30 characters each) and up to 4 descriptions (90 characters each), and Google tests combinations to find what works.

• Use the searcher’s words. If the ad group is about alternatives, the headlines should say ‘alternative’.
• Mix your headlines. Include the benefit, the proof (customers, results, ratings), the differentiator, and a clear call to action.
• Pin sparingly. You can pin a headline to a position, but every pin reduces Google’s ability to test. Pin only what must always show.
• Add assets. Sitelinks, callouts, and structured snippets make your ad bigger and give people more reasons to click.

Bidding: start simple

1. Weeks 1–3: Start with Maximize clicks with a maximum cost-per-click limit, or Manual CPC. The goal is to collect data at a controlled cost.
2. Once conversions are coming in: Switch to Maximize conversions so Google starts optimising for signups instead of clicks.
3. Once cost per conversion is stable: Add a target CPA, set close to what you are actually achieving, then lower it gradually towards your target.

What about AI Max?

AI Max is Google’s AI layer for Search campaigns. It expands which searches you match, adjusts your ad text, and can choose which landing page to send people to. During 2026, Google is upgrading its older Dynamic Search Ads into AI Max, so you will see it in your account.

For a first test, keep AI Max off until you have conversion data and know what ‘good’ looks like. Then turn it on as a deliberate experiment and compare the results. Otherwise, you won’t be able to tell whether a result came from your keywords or from Google’s expansion.

Location, language, and schedule

Target only the countries or regions you actually sell to, and the languages your customers use. If you sell B2B and your team only responds on weekdays, it’s worth checking later whether weekend clicks convert. Don’t restrict the schedule on day one; let the data tell you.

5. Landing Pages That Convert

The ad gets the click. The landing page gets the conversion. You can have perfect keywords and great ads and still fail the test because of the page people land on.

• Match the message. The headline on the page should echo the keyword and the ad. If someone searched for ‘[competitor] alternative’, the page should say so in the first line.
• One page, one goal. One clear call to action. Remove anything that pulls people away, including the full site navigation if you can.
• Show proof early. Customer logos, a short testimonial, or a result, near the top of the page.
• Answer objections. Pricing, integrations, security, setup time. Whatever stops people from signing up, answer it on the page.
• Make it fast on mobile. A slow page wastes the click you just paid for.
• Match the page to the intent. Alternatives ads go to a comparison page. Category ads go to a product page. Don’t send ads to your homepage.

If you are building SEO content at the same time, your comparison and alternatives pages can double as ad landing pages. You get to test their conversion rate with paid traffic long before they rank.

6. Tracking What Works (and What Doesn’t)

You don’t need expensive tools to run a test. Google Ads and GA4 give you everything you need at this stage. What you need is a routine.

A weekly routine

1. Search terms report: Look at the actual searches that triggered your ads. Add irrelevant ones as negative keywords. Add the ones that convert as new exact match keywords.
2. Keywords: Pause keywords that have spent two to three times your target cost per conversion with no conversions.
3. Ads: Check which headlines and descriptions are performing, and replace the weakest ones.
4. Landing pages: Check the conversion rate of each page. If clicks are coming in and nobody converts, the problem is usually the page, not the ad.
5. Budget: Check that spend is on pace, and that your best ad groups aren’t running out of budget early in the day.

The numbers that matter

Judge your results against your target, not against Google’s optimisation score. The optimisation score measures how many of Google’s recommendations you have adopted, not whether your ads are making money. The same goes for Quality Score: it’s a useful diagnostic, not a goal.

7. What Not to Waste Money On (Yet)

• Performance Max: It runs your ads across all of Google’s surfaces, with less control over where they show. It needs conversion data to work well. Try it after search is proven, not before.
• Display and YouTube: These are awareness channels. They have their place, but not in a test that’s meant to tell you whether people searching for your category will buy.
• Broad match on a new account: Without conversion data, broad match mostly finds you expensive, loosely related clicks.
• Accepting every recommendation: Google’s recommendations and account reps can be helpful, but their suggestions tend to push towards more spend and more automation. Weigh each one against your target.
• Too many campaigns: One well-run campaign teaches you more than five underfunded ones.
• Changing things every day: Every significant change resets part of Google’s learning. Make a change, then give it at least a week before you judge it.

8. Wrapping Up: Your First 90-Day Google Ads Test

Days 1–30: Research, setup, and launch

• Set your target cost per conversion and work out the budget you need.
• Research keywords, starting at the bottom of the funnel, and build your negative keyword list.
• Set up and test conversion tracking.
• Get your landing pages ready, one per intent.
• Build one Search campaign with three to five ad groups and launch it.

Days 31–60: Optimise every week

• Run the weekly routine: search terms, keywords, ads, landing pages, budget.
• Switch to Maximize conversions once conversions are coming in.
• Test new ad copy and landing page changes, one at a time.

Days 61–90: Reach a verdict

Compare your cost per conversion against the target you set on day one. Then decide:

• Scale: You are at or near target. Increase budget, expand keywords, and test AI Max or Performance Max deliberately. This is also the right time to hand it to an in-house marketer or a specialist agency.
• Fix: You are close, and there is a specific, fixable reason: a weak landing page, the wrong offer, or tracking gaps. Fix it and run another cycle.
• Stop: You are far from target with no clear fix. Pause, move the budget to another channel, and keep the search terms data. It’s one of the best inputs you will ever get for your SEO.

Whatever the outcome, you will end the 90 days knowing something most founders only guess at: whether people searching for what you do will pay for it, and what it costs to reach them.