How much should a small business spend on ads?

Most advice here is a percentage of revenue, which tells you nothing useful. Here is a way to get to a number you can defend.
Start from what a customer is worth
Work out two figures before you spend anything.
What a customer is worth over their whole relationship with you, not just the first sale. A gardener with a client who stays three years is not selling one job.
What proportion of that you are willing to pay to acquire them. Twenty to thirty percent is a common comfortable range for a service business; it leaves margin for delivery and profit.
Multiply them. That is your maximum cost per customer. Everything else is arithmetic.
Then work backwards
If a customer is worth 1,000 dollars and you will pay 25 percent, you can spend 250 dollars to get one.
If one in ten enquiries becomes a customer, you can spend 25 dollars per enquiry.
If one in twenty clicks becomes an enquiry, you can spend about 1.25 dollars a click.
Now you can look at what clicks actually cost in your market and know immediately whether this is viable. If clicks cost 4 dollars, the maths does not work yet, and more budget will not fix it. Either the conversion rate has to improve or the customer value has to rise.
The learning budget
Before scaling, you are not buying customers, you are buying answers. Which message, which audience, which offer.
Budget for roughly fifty to a hundred conversions worth of spend before you judge anything. Below that you are reading noise, and most people stop a campaign that was working or scale one that was lucky.
For many small businesses that is a few hundred to a couple of thousand dollars. If you cannot afford the learning budget, you cannot afford to run ads properly, and earned traffic is the better place to start.
What quietly eats the budget
Sending everyone to your homepage. The ad promised something specific; the homepage answers something general. Send them to the page about the thing they clicked on.
Running three messages at once with no way to tell which worked.
Pausing and restarting constantly. Every restart throws away what the platform learned about who converts.
Having no tracking of what actually produced a sale, so you optimise for clicks instead of customers.
Paying for your own brand name when you already rank first for it organically, which is sometimes defensive and often just a tax.
When to stop
Stop when the cost per customer is above the number you calculated and three honest attempts to fix it have failed.
That is not failure, it is information: this channel does not work at your current margins. It may work later when a customer is worth more or your page converts better.
The part that is not about budget
An ad can only be as good as what happens after the click. If the page is slow, vague, or does not answer the question that made them click, no budget rescues it.
The cheapest improvement to any ad account is usually not in the ad account.
Keep reading
Digital ProductsHow to price a digital productWhy cost-based pricing fails for files, how to price against the outcome instead, and what to do when nobody is buying at any price.
Digital ProductsWhat you need in place before selling digital productsThe practical checklist before your first sale: delivery, refunds, tax, licensing and the file itself, and which parts genuinely matter at the start.
SEOSEO or paid ads: which should you do first?The honest trade-off between search and paid traffic for a small business, what each actually costs, and the case where doing either one first is a mistake.
Comments