THE SHORT ANSWER There is no honest fixed number or percentage. Your budget is the output of three inputs: what a customer is worth in margin, what you can afford to pay to acquire one, and what each channel actually charges you per closed customer. Answer those from your own books and CRM, fix conversion leaks first, and the budget writes itself.
Every week an owner asks us some version of the same question: how much should I be spending on marketing? And every week the internet answers with the same folklore: small businesses should spend some fixed percent of revenue. New businesses more. Established ones less. It sounds like math. It's astrology.
This is the budgeting method we actually use — built from your own numbers, not from an average of businesses that have nothing in common with yours.
Why "percent of revenue" is astrology
Those percentages are averages across companies that share nothing but a tax form. A software firm with fat gross margins, a restaurant running on thin ones, a law firm where one client pays for a year of ads — blend them together and you get a number that describes none of them.
Three numbers decide what you can actually afford to spend, and not one of them appears in the percent-of-revenue rule:
- Your margin per customer. If yours is half the average's, the "standard" percentage quietly overspends you into a hole.
- Your close rate. Two businesses buying identical leads at identical prices have completely different real marketing costs if one closes twice as often.
- Your retention. A customer who renews for years can justify an acquisition price that would bankrupt a one-and-done business.
The tell: the rule hands the same number to a personal injury firm and the pizzeria downstairs from it. Any rule that can't tell those two businesses apart isn't a rule. It's a horoscope with a spreadsheet attached.
Budget from your own unit math instead
A real budget is the answer to three questions, all answerable from your own books and your own CRM.
- What is a customer worth? Margin, not revenue, multiplied by how long they actually stay. A big sale at a thin margin from a customer who never returns is worth a fraction of its sticker price. We walk the full arithmetic in our guide to CAC and LTV.
- What can you afford to pay for one? That's your allowable CAC: the fraction of customer worth you'll spend to acquire a customer, at a payback speed your cash flow survives. You set it deliberately. It is the spine of the whole budget.
- What does each channel actually charge you per closed customer? From your CRM — not the platform dashboard. The dashboard sells you cost per click and cost per lead. Those are ingredients, not prices. The price is spend divided by customers closed, per channel, from your own records.
Then notice what the biggest lever in this math is — because it isn't in the ad account. If you close 20 percent of your leads and you get to 30, the same leads and the same spend produce 50 percent more revenue. That one move cuts the real cost of every channel you run, simultaneously, without renegotiating a single click. Budgeting and conversion are the same conversation.
The NYC layer: the auction doesn't care about averages
Everything above applies anywhere. New York adds one brutal amplifier: this city's ad auctions are fiercely contested, and contested auctions punish weak conversion hard.
Here's what that looks like in Semrush data we pulled in July 2026:
| Search auction, NYC market | Cost per click |
|---|---|
| Personal injury, New York City | $73.56 |
| "dentist staten island" | $2.79 |
Semrush, pulled July 2026. Auction prices move — treat these as a dated snapshot, not gospel.
The spread is the point. Same city, same month, and one click costs 26 times the other. A "citywide average CPC" blending those two numbers describes nobody — which is why your category and your borough decide your math, not the market average. At $73.56 a click, a slow callback is a luxury purchase; at $2.79, you can afford to test and learn. Either way, the budget still comes from your unit math — the auction just decides how expensive your mistakes are.
We publish the full picture across the region in our tri-state search market report, and cover how we work with businesses here on our New York page.
The sequencing rule: fix conversion before you scale spend
Most owners set budget by asking "how much more should we spend?" The prior question is: how much of what you already buy is leaking? Leads called back hours later. Follow-up that stops after one attempt. First calls that open with "so, uh, you filled out a form?"
The cheapest budget increase available to any New York business is a faster callback. It costs nothing, and it reprices every lead you already pay for. Scaling spend into a leaky funnel doesn't grow the business — it buys more leaks, at New York auction prices.
The build order for a first NYC budget
- Put customer worth on paper. Margin per customer times how long they stay. One number you can defend to a skeptic.
- Set your allowable CAC. Decide what fraction of that worth you'll pay for a customer, and how fast it must pay back given your cash position. Write it down. It's the line every channel gets measured against.
- Audit your conversion machine before spending a dollar. Time-stamp your speed to lead, count your follow-up touches, listen to a first call. Fix what's broken first — it's the cheapest capacity you will ever add.
- Price your channels from the CRM. Last 90 days: spend per channel divided by customers closed per channel. Fund what closes customers under your allowable CAC. Starve what doesn't. Ignore the platform dashboard's opinion of itself.
- Let the budget be the output. Your number is whatever your under-CAC channels can absorb while conversion holds. Re-run the math monthly. Scale winners, cut losers, no ceremony.
When to hire help — and when to do it yourself
Do it yourself if you run one main channel, you can pull cost-per-closed-customer from your own CRM, and your close rate is healthy. The framework above is the whole job; it takes an afternoon and a spreadsheet.
Get help when any of three things is true: you can't produce per-channel cost per closed customer (that's a tracking problem, and it must be fixed before another dollar goes out); you're bidding in a contested category where each mistake costs more than dinner; or the real leak is your close rate — which is a sales problem wearing a marketing costume, and no ad budget fixes it. That third case is most of what our New York sales consulting work turns out to be.
From our own floor: we never budgeted from a percentage. FreedInsure started in February 2023 — two people, zero startup capital, $227 in revenue the first month. Every marketing dollar after that went where the unit math said it should: we built our own lead engine and bought vendor leads, and funded each one by its cost per closed customer, nothing else. That book now averages $1.3M a year with roughly three-quarters of clients renewing. The math worked because we let it make the decisions.
Straight answers
There's no honest fixed number — and no honest percentage either. The budget is an output of three inputs: what a customer is worth to you in margin, what you can afford to pay to acquire one, and what each channel actually charges you per closed customer in your category and your borough. Answer those from your own books and CRM and the budget writes itself. Anyone who hands you a number before asking those questions is reciting folklore.
Good is a ratio, not a number. A CAC is good if the lifetime margin a customer generates pays it back several times over, fast enough that your cash flow survives the wait. The identical acquisition cost can be a bargain for a business whose customers renew for years and a disaster for one that sells once. Compare your CAC to your own LTV — never to a benchmark from a business that isn't yours.
Because ad prices are set by auction, and New York's auctions are fiercely contested — a dense market of businesses bidding for the same customers. Auction prices climb to whatever the winner's economics can bear, so categories with enormous case values bid clicks to extreme levels while lower-stakes local categories stay cheap a borough away. The city doesn't have one price. Your category, your borough, and your conversion rate decide yours.
Not into new clicks. The first dollar goes into converting the demand you already have: a faster callback on every lead, a follow-up cadence that outlasts the prospect's indecision, and a planned first call instead of an improvised one. Those fixes reprice every lead you're already paying for. Buying more traffic before fixing conversion just buys leaks at New York prices.
