“How much should I spend on marketing?” is really three questions wearing one trench coat: how much can I afford, where should it go, and how will I know it worked? The internet answers with a percentage — “spend 7–8% of revenue!” — which is a bit like prescribing calories without asking whether the patient is bulking or recovering.
Here’s the fuller answer: the percentage bands and when they’re honest, the stage-based version that’s actually useful, and the measurement habit that makes any budget defensible.
The percentage rules — and what they’re for
The commonly cited bands for local service businesses:
- Maintaining: 3–6% of revenue — established, referral-fed, protecting position
- Growing: 7–12% — building visibility, feeding crews, entering seasons
- Aggressive/new: 10–15%+ — no book yet, buying speed while foundations mature
These are sanity rails, not prescriptions. Their real use is catching the two classic errors: the established business spending 0% (“we’ve never needed marketing” — until the founder’s referral network retires), and the new one spending 20% on ads while its Google profile sits half-empty and its phone rings out.
The sequence matters more than the amount
Local marketing has a strict order of operations, because the free assets multiply every paid dollar spent after them:
Layer 1 — the free foundations (time, not money): a complete Google Business Profile, a fast site that converts, and the review habit. These compound forever and set the conversion rate for everything downstream.
Layer 2 — the multipliers (small money): missed-call text-back, quote templates, follow-up cadence — the systems that stop paid leads from leaking.
Layer 3 — paid speed (real money): LSAs, search ads, social — rented visibility that works exactly as well as layers 1 and 2 allow.
The most common local budget mistake isn’t the amount — it’s buying layer 3 before layers 1 and 2 exist. A $1,000/month ad budget pointed at a weak profile and a slow site performs like $300; the same spend after the foundations performs like $2,000. When owners say “ads don’t work in my area,” the audit usually says otherwise — run yours free before concluding anything.
Budgets by stage, concretely
Brand new (first 100 customers): mostly time. Foundations built by hand, the first-customers playbook worked hard, and — once something converts — one paid channel at $300–800/month. One. Channel-splitting a small budget guarantees five sets of noise and zero signal.
Growing (filling crews/routes): 7–12% of target (not current) revenue, weighted to your trade’s buying season — a landscaper’s annual budget belongs mostly in February–April, a gutter company’s around storm and leaf season. Flat monthly spend through a seasonal demand curve is quiet waste.
Established (referrals + rankings carrying): 3–6%, increasingly weighted to defense — the review base, the site, the profile — plus ads as a thermostat: on when churn ticks up or capacity opens, down when booked solid. The foundations never go to zero; they’re why the percentage got small.
The measurement habit that justifies any budget
One spreadsheet, five columns, monthly: channel → spend → leads → booked jobs → cost per booked job. Attribution at local scale is unglamorous and effective: ask every new customer “how’d you find us?” and write it down.
Three rules fall out of the sheet:
- A channel gets 90 days of honest effort before judgment — local data is noisy weekly, clear quarterly
- Judge on booked jobs, never clicks — the cheapest leads are frequently the worst ones
- Feed winners before adding channels — doubling a working channel beats debuting a new one, almost always
And count lifetime value where it applies: a recurring-service client or a contract customer is worth years, not tickets — budgets that look expensive per job are often cheap per relationship.
The bottom line
Spend time before money; foundations before ads; one channel before five. Hold the percentage rails loosely (3–6% maintaining, 7–12% growing, more when new), weight spend to your season, and let a five-column spreadsheet — not a feeling — decide what continues.
And if the honest answer is that layer 1 never quite gets built between jobs: that’s the exact gap we sell into — the profile, the site, the reviews, maintained monthly from $29, so every dollar you spend above it actually multiplies.