How much should I spend on Google Ads before deciding if it works?

Judge by Conversions, Not by a Dollar Amount

There's no fixed dollar figure that tells you Google Ads works — the honest answer is a conversion count, not a budget. You need enough recorded conversions to separate a real signal from random luck, and the working number is roughly 30–50 conversions before you draw a conclusion.

Here's why. If you spend $1,000, get two leads, and one becomes a customer, you genuinely cannot tell whether the channel is profitable or you got lucky on a single click. Flip a coin five times and you might get four heads — that doesn't make it a weighted coin. Paid search behaves the same way: small samples produce wild swings in cost per lead that have nothing to do with the underlying performance. Thirty-plus conversions is where the numbers start to stabilize enough to trust.

This reframes the whole question. Instead of asking 'how much should I spend,' ask 'what does it cost to buy 30–50 conversions in my category?' That answer depends on two things you can estimate up front: your cost per click and your landing-page conversion rate. Multiply them out and you get a real test budget tied to your business — not a number borrowed from a blog post about a company in a different industry, a different city, and a different price point. A campaign isn't a failure at $800; it's simply unjudged.

The Math: Turning a Conversion Target Into a Budget

To translate '30–50 conversions' into a budget, you only need two estimates: your cost per lead and how many leads you want to see. Multiply your expected cost per lead by 30–50, and that's your honest test budget.

Work it backwards. If clicks in your category cost $3 and a well-built landing page converts 5% of visitors, every conversion costs about $60 in ad spend ($3 ÷ 0.05). To collect 40 conversions, you need around $2,400 — call it $1,500–$3,000 across the test, depending on how clean your tracking and pages are. That's a typical small-business range in a low-to-mid CPC category, and it usually plays out over two to three months rather than all at once.

Now change one input. In a high-value category — legal, insurance, some home services — clicks run $12 or more, so the same 5% conversion rate puts each lead near $240, and a 40-conversion test costs several thousand dollars more. Same logic, very different number. This is exactly why a $2,000 'try it for a month' test fails in those markets: it doesn't buy enough clicks to ever reach a verdict, so people conclude 'Google Ads doesn't work' when the truth is 'we never spent enough to find out.'

If you don't know your numbers yet, use conservative placeholders, run for the window below, then recalculate with your real data. The first round's job is often just to discover your true cost per lead.

Give It Time, Not Just Money

Budget alone isn't the test — time is the other half, and rushing it is the most common way owners misread a campaign. Plan on at least 60–90 days before a final verdict, even if you hit your conversion count sooner.

The first two to four weeks are a learning period by design. Google's automated bidding needs conversion data to calibrate, and during that window it deliberately explores — testing audiences, placements, and bids to find what works. Results in those early weeks are volatile and usually worse than the account's eventual steady state. Judging Google Ads on week one is like rating a new hire on day two: you're measuring the ramp, not the performance.

There are seasonal and sales-cycle reasons to wait too. If your business has busy and slow months, a 30-day test that lands in a slow patch tells you very little. And if your sales cycle is long — common in B2B and high-ticket services — a lead generated in week three may not close until month two or three, so cutting the test early hides the customers it actually produced.

That said, 60–90 days isn't a blank cheque. You should see direction within the first month: are conversions trending in, is cost per lead improving as the account learns, is search-term data showing relevant queries? Stable-but-bad after a fair window is a real answer. Volatile-but-improving means keep going. The discipline is committing to the window before you start, so a slow week doesn't trigger a premature shutdown.

Make Sure You're Testing the Right Thing

Before you spend a dollar judging Google Ads, make sure the test itself is valid — otherwise you'll spend your whole budget measuring the wrong thing. The most common reason a 'fair test' gives a false verdict is broken conversion tracking.

If calls, form fills, and sales aren't tracked accurately, every conclusion you draw is guesswork. Plenty of accounts that 'don't work' are actually generating customers the data never recorded — phone calls that weren't imported, a thank-you page that never fired a tag, offline sales that closed in your CRM and never made it back to Google. You're judging a channel while half the scoreboard is dark. Confirm conversion tracking is firing correctly before the test starts, not after it disappoints.

The other thing that invalidates a test is sending paid clicks somewhere they can't convert. If your ads point at a slow homepage instead of a focused landing page built for that search, your conversion rate craters and your cost per lead looks terrible — but that's a landing-page failure wearing a Google Ads costume. Tight keyword-to-page relevance, fast load, and a clear single action are part of the test, not optional extras.

Finally, decide your verdict metric before you start, and make it cost per acquired customer — not clicks, impressions, or even raw leads. Cheap leads that never answer the phone aren't a win. At SearchPod we'll tell you up front roughly what a fair test costs in your category, make sure tracking is right before spending, and run it month-to-month so you're never locked into a test that was doomed by its own setup.

Related questions