Your First Ad Budget: How to Spend It So You Learn Something
The most common way small businesses waste their first ad budget is not by choosing the wrong platform or writing bad copy. It is by spending too little, across too many things, for too short a time, without the tracking needed to learn anything from it. Then concluding that ads do not work for their industry.
A first budget has a different job from an ongoing one. It is not there to be profitable. It is there to buy you a reliable answer to one question: what does a customer from this channel cost? Once you know that, every later decision is arithmetic.
Work Backwards From What You Can Afford
Before choosing a number, work out what a customer is worth to you, because that sets the ceiling on everything else.
- Average customer value. Not the first invoice, the whole relationship. A gym member worth sixty a month who stays fourteen months is worth eight hundred and forty, not sixty.
- Gross margin. What is left after delivering the work. If you keep forty percent of that eight hundred and forty, you have three hundred and thirty-six to play with.
- What you are willing to pay to acquire one. A common starting point is somewhere between a fifth and a third of gross margin. Say a hundred.
- Your enquiry-to-customer rate. If one in four enquiries becomes a customer, you can afford about twenty-five per enquiry.
That last figure is your target cost per lead, and it is the number the whole test is measured against. Without it you are looking at ad platform metrics with no idea whether they are good or bad.
Sizing The Test
Here is the rule that saves the most money: your test budget should be enough to produce at least thirty leads. Below that, the numbers bounce around too much to mean anything, and you will be making a permanent decision on a random fortnight.
Thirty leads at a target of twenty-five each is seven hundred and fifty. Round up, because your first attempt will be worse than your fourth, and budget around a thousand for the test. Spread over four to six weeks rather than blown in ten days, because the platforms need time to settle and because you need to see more than one week's worth of buying behaviour.
If a thousand is genuinely beyond reach, the honest answer is to fix your website and follow-up first and come back to ads when the budget exists. Underfunding a test does not produce a cheap answer. It produces no answer.
One Platform, One Audience, One Offer
The instinct is to spread the budget so nothing is missed. Do the opposite. A thousand pounds split across Google, Meta, LinkedIn and TikTok is four tests that all fail to reach significance. A thousand on one platform is one test that gives you a real answer.
Choosing between the two obvious candidates:
- Google Search if people already search for what you sell. Emergency plumbers, accountants, solicitors, dentists, anything people look up when they need it. You are meeting existing demand, which is why intent is high and clicks are expensive.
- Meta if people would want it but would not think to search for it, or if you can define the audience by who they are rather than what they typed. Consumer products, local classes, lifestyle services, anything visual. Cheaper attention, colder audience, more dependent on the creative.
Within the platform, keep it equally tight. One clearly defined audience or a small set of high-intent keywords. One offer. Two or three ad variations testing the message, not the layout.
Spread thin, you learn nothing and blame the platform. Concentrated, you get an answer you can act on, even when the answer is no.
The Landing Page Is Half The Test
Sending paid traffic to your homepage is the most common and most expensive mistake in this whole process. The homepage is built for everybody. Your ad promised something specific, and the visitor arrives looking for it.
Build one page per offer. It should carry the same headline as the ad, one action, no site navigation to wander off into, evidence near the top, and a form short enough to finish on a phone in under a minute. Every field you add costs you completions, so ask for what you need to make contact and get the rest on the call.
Tracking, Or The Whole Thing Is Theatre
Set this up before a single ad goes live. It is not optional and it takes an afternoon.
- Conversion tracking firing on the actual enquiry, not on a page visit or a button hover.
- Call tracking if the phone matters. A dedicated number on the landing page tells you which calls came from ads and which did not.
- Source tagging carried into your CRM so a closed sale still remembers it started as an ad click.
- Outcome recording. Someone has to mark leads as won or lost. Without it you will optimise toward cheap leads that never buy, which is worse than not advertising.
Judging The Result Without Kidding Yourself
At the end of the test, ignore impressions, ignore click-through rate, ignore engagement. Three numbers matter: leads generated, cost per lead, and how many became customers.
Then one of four things is true:
- Cost per lead is under target and they convert. Increase budget by about half, monitor, repeat. Do not double overnight.
- Cost per lead is fine but nobody buys. The traffic is wrong or the leads are unqualified. Tighten targeting or add a qualifying question to the form.
- Cost per lead is too high but the leads that do come are excellent. Usually a landing page or offer problem rather than a traffic problem. Fix the page before touching the targeting.
- Nothing worked. This is a legitimate outcome and worth the money. You now know this channel is not where your next customers are, which stops you spending another year wondering.
The Bottom Line
Treat the first budget as tuition rather than turnover. Concentrate it, run it long enough to mean something, measure it properly, and be willing to accept an unwelcome answer. Businesses that do this find their number within six weeks and scale from a position of knowledge. Businesses that dabble spend the same money over two years and still cannot tell you what an ad-sourced customer costs.