What a marketing budget
should actually be

Why the percentage rules are useless to you

Percentage-of-revenue guidance averages across industries with wildly different margins, customer values and sales cycles. A dental implant practice and a pizza restaurant cannot share a rule. Neither can two dental practices where one closes 60% of consultations and the other closes 20%.

Worse, it reasons backward. It sets spend from what you happen to have rather than from what a customer is worth. Two businesses of identical size should spend very differently if one earns $400 per customer and the other $4,000.

The four numbers that actually decide it

You need these, and most owners can produce them in an hour:

• Average value of a customer. Not one transaction — what they are worth over the relationship, including repeat work and referrals.
• Lead-to-customer close rate. Of the enquiries you get, what fraction become customers. Be honest; most people overstate this.
• Cost per lead. What you currently pay to generate one enquiry, by channel.
• Capacity. How many additional customers you can actually serve well this quarter.

Two of those give you the ceiling immediately. Customer value multiplied by close rate is the absolute maximum you can pay for a lead and break even. If a customer is worth $2,500 and you close 30%, a lead is worth $750 at break-even. Paying $85 for that lead is not expensive; it is an 8.8× return before you have negotiated anything.

That is the calculation the ROI calculator on our results page runs, and it is worth doing with your own numbers before you read another agency proposal.

Sanity-check your cost per lead against your category

LocaliQ’s 2026 benchmarks put average cost per lead from paid search at $66.69 across all industries — and the spread by category is wide. Automotive repair around $29.96, physicians $40.04, health and fitness $67.36, dentists $72.97, home improvement $90.92, real estate $102.51, attorneys $131.63.

If your figure is close to your category, your acquisition is roughly normal and growth comes from raising close rate or customer value. If it is far above, something is broken upstream — usually tracking, match types or the landing page. If it is far below, verify what you are counting as a lead.

Capacity is the constraint nobody plans for

The most common failure of a successful campaign is not that it fails. It is that it works, the phone rings more than the business can answer, and half the value leaks out through missed calls and slow follow-up.

Before increasing spend, ask what happens at double the current enquiry volume. Who answers at 4pm on a Friday? How fast does a web form get a reply? If the answer is “we would manage,” the honest move is to fix answering before adding budget. It is cheaper and the return is immediate.

Timing: not every channel pays back on the same schedule

Paid search produces leads in week one and stops the day you stop paying. SEO produces nothing for months and then compounds. Ahrefs, tracking two million pages over a year, found only 5.7% reached the top ten for even one keyword within twelve months — and of the ones that did, most took roughly two to six months.

That has a direct budget implication. If you need customers this quarter, weight paid. If you are building an asset for the next three years, weight organic and accept a period of paying for nothing visible. Most businesses need both, in a ratio that shifts as the organic work matures.

A defensible starting point

Work out what a customer is worth and what fraction you close. Multiply for your break-even cost per lead. Decide what return you need — three times is a reasonable floor for a business with normal margins. Divide to get your target cost per lead, and check it against your category benchmark. Multiply by how many customers you can genuinely serve.

That is your budget. It is defensible, it is tied to your own economics, and when someone proposes a number you can immediately say whether it makes sense. Businesses growing fast in markets like Allen and Frisco tend to be the ones that did this arithmetic before they were sold a package, not after.