How Much Should a Contractor Spend on Marketing? We Read 9 Guides. The Number They All Quote Comes From a Dead Page.

It is October. Somewhere between the last roof of the season and the first snow, most contractors sit down and pick next year's marketing number. The usual way to pick it is to search "how much should a contractor spend on marketing," read the first guide, and write down 7 to 8 percent of revenue.

We read the first 9 guides that come up for that search. 4 of them quote the same rule: "the SBA recommends 7 to 8 percent of revenue for businesses under $5 million." None of the 4 link to it. We went looking for the page. It is gone. The address it traces to on sba.gov returns a 404, and we could not find an archived copy. The one marketing-budget post the SBA still publishes says, in its own words, there is "no hard and fast answer."

So this post does two things. First, it shows you where the number everyone quotes actually comes from, and why it says very little about a two-crew roofing company. Second, it builds a 2027 budget the other way round: from the jobs you want, your close rate, and the price Google itself publishes for a click in your trade. We pulled that bid data on October 5, 2026, for 15 trades, Canada-wide and US-wide. The percentage falls out at the end, and it is a different number for every trade.

Nine Guides, One Dead Link: Where the 7 to 8 Percent Rule Comes From

Here is what each guide tells a contractor to spend, what it cites for that, and whether it shows any working. The ranges are the guides' own words, so the inconsistency is theirs, not ours.

Three things stand out. The "growth mode" advice runs from 8 to 12 percent in five guides to 12 to 18 percent in one. The "established" advice runs from 3 to 8 percent in one guide to 10 to 15 percent in another. Take a contractor doing $1 million a year and the same guides will tell them to spend anywhere from $30,000 to $150,000. That is not a benchmark. That is a coin flip with extra steps.

Second, 4 of the 9 lean on the SBA for the 7 to 8 percent figure and none of them link to it, because there is nothing to link to. The page is dead. The SBA's live post on the subject, written by Rieva Lesonsky in July 2019, does not contain "7 to 8 percent" or "$5 million" anywhere. It says there is no hard and fast answer, then quotes The CMO Survey's 2018 average of 7.9 percent of revenue. Somewhere in the last decade the "about 8 percent, from a survey of big companies" became "the SBA recommends 7 to 8 percent for small businesses," and the copies have outlived the source.

Third, and to be fair to the guides, 7 of the 9 do show some backwards math, usually a line like "100 leads at $150 each is $15,000." That is the right idea. What none of them do is tell you where the $150 came from, or that in your trade the real number might be $60 or $400. That is the gap we fill below.

A Benchmark From Billion-Dollar CMOs Is Not a Budget for a Two-Crew Roofer

The two surveys that the honest guides cite are real, current, and well run. They are also measuring something you are not.

Gartner's 2026 CMO Spend Survey puts marketing at 7.8 percent of company revenue, up from 7.7 percent in 2025. It surveyed 401 marketing leaders, the majority at companies with more than $1 billion in annual revenue, between January and March 2026. The CMO Survey from Duke, Deloitte and the American Marketing Association puts it at 9.0 percent of revenue in its 2026 report, the lowest reading since 2021, from 308 US marketing leaders, 97 percent of them vice-president level or above. Business-to-consumer service companies in that survey, the closest bucket to a contractor, sit at 7.2 percent. Mining and construction together make up 1.3 percent of the sample, which is about four people, and they report the lowest marketing share of any sector at 0.3 percent of revenue.

So the "7 to 8 percent" that survives in the guides is a weighted average of what Fortune-scale marketing departments spend on brand, media, agencies, software and staff, across every industry at once. It was never a recommendation for anybody, let alone a $900,000 plumbing company in a two-town service area. Spending 7.8 percent because a bank and a software company average 7.8 percent is like pricing a roof off the average house price in the country.

Percent of revenue is a fine way to report a budget after the fact. It is a poor way to set one, because it hides the only three things that actually decide the number: how many jobs you want, how many leads it takes to book one, and what a lead costs in your trade.

Build It Backwards: Jobs Needed, Close Rate, and Google's Own Bid Range for Your Trade

The formula is short. Decide the new revenue you want next year. Divide by your average ticket to get jobs. Divide by your close rate to get leads. Multiply by cost per lead. That is the paid part of your budget. Everything else, the website, the Google Business Profile, reviews, a CRM, is a fixed cost you add on top, and we come back to that at the end.

The only input most contractors cannot fill in is cost per lead, so the guides make one up. We would rather use Google's number. The Google Ads Keyword Planner publishes, for every search, the "low" and "high" top-of-page bid: the 20th and 80th percentile of what advertisers actually paid for a top-of-page click over the last year. We pulled it on October 5, 2026, for 54 searches across 15 trades ("roof repair," "plumber near me," "furnace replacement," and so on), once for Canada as a whole and once for the United States, covering September 2025 to August 2026. Both are in Canadian dollars because that is our account currency, so the US column is a CAD figure, not a USD one.

To turn a click into a lead you need a conversion rate. The best public figure we know of is LocaliQ's 2026 search advertising benchmarks, updated June 2026 from thousands of accounts: home and home improvement campaigns convert 8.05 percent of clicks into a call or form. So the lead ranges below are simply each bid divided by 0.0805. Treat them as rough. A tight campaign with a good landing page beats that rate; a sloppy one does worse.

Two things to notice before you plug in. The spread inside a single trade is roughly eight to one: a Canadian plumber can pay $4.53 or $36.53 for the same kind of click depending on the keyword, the town and how well the account is run. That spread is why a guide's single "$150 per lead" is not a number you can budget on. And the US market, in Canadian dollars, runs about 2.4 times Canada's on the high bid, median across the 15 trades. A US roofer and a Canadian roofer reading the same guide are in different businesses. If you are in the Okanagan, we published the Kelowna and Vernon numbers last week; they sit close to the Canada column.

Three Worked Budgets: Roofer, HVAC, Plumber, and Why the Percentage Comes Out Different Every Time

Here is the formula run for three trades, each chasing the same $500,000 of new revenue in 2027, using the Canada column. The tickets and close rates are example assumptions, not survey data. Swap in yours; the point is what happens to the percentage, not the exact dollars.

The roofer needs 42 roofs, so about 139 leads. Even at the top of Google's bid range the paid budget is $46,944, or 9.4% of the new revenue, and at the low end it is under 2 percent. Each booked roof costs $213 to $1,127 to acquire on a $12,000 ticket. A roofer who reads "8 percent" and spends $40,000 is probably fine; the ticket is big enough to carry almost any click price. The real risk for the roofer is not overspending, it is having 139 leads land and only answering half of them.

The HVAC company needs 77 installs and 220 leads. At the low end of the bid range the budget is 3.1% of the target; at the high end it is 19.5%. That is the same company, the same goal, and a 6x difference in spend depending on whether its account is buying cheap clicks or expensive ones. "Percent of revenue" cannot tell you which one you are. Your search terms report can.

The plumber is where the rule breaks. To book $500,000 of $450 service calls you need 1,111 jobs and about 2,222 leads, and Google's bid range puts that at $124,444 to $1,008,889. The paid budget is 24.9% of the revenue at best and more than all of it at worst. Each booked call costs $112 to $908 to win on a $450 ticket. No percentage fixes that. The only way the arithmetic works for a low-ticket trade is if the first call is not the last one: the customer comes back, leaves a review, refers a neighbour, or signs up for a maintenance plan. For a plumber, the budget question is really a lifetime-value question, and paid search is the door, not the business.

So what should the number be?

Run the formula with your own ticket, close rate and the bid row for your trade, and you will get a range. Then apply three checks before you write it into the plan.

1. Can the business cash-flow it? The budget lands before the jobs do. If a $97,363 year of ads means missing payroll in March, the right budget is the one you can fund through the slow quarter, and the goal gets smaller to match. Front-load it into the months your trade actually gets searched; we mapped those for 18 Okanagan trades in our search calendar, and the same logic applies anywhere with a real winter.
2. Can the crew do the work? 42 extra roofs is a second crew. 1,111 extra plumbing calls is three more vans. Marketing that books work you cannot do produces one-star reviews, which costs you the next year's leads for free.
3. Are the owned assets already paid for? The paid budget above assumes the click lands on a page that converts, a Google Business Profile with recent reviews, and a phone that gets answered. If any of those is missing, money spent on clicks leaks out before it becomes a lead. A website that converts and a Google Business Profile that ranks are not a share of the marketing budget; they are the floor it stands on, and they keep paying after the ads stop.

Where that leaves most contractors we work with: owned assets first, at a fixed cost you can name; then a paid budget derived from the formula, not from a percentage; then a monthly review of cost per booked job, not cost per lead, because cost per lead is lying to you and it has risen sharply for many trades anyway. If the formula says paid search cannot carry your trade on its own, that is not a failure. It is the data telling you that referrals, repeat work and reviews have to do part of the job, and you can budget for those too.

We sell Google Ads management, so read our preference for owned assets first as the view of a company that gets paid either way, and weigh it accordingly. The bid data is Google's. The surveys are Gartner's and Duke's. The 7 to 8 percent is nobody's.

How we did this: on October 5, 2026 we fetched the first 9 guides returned for "how much should contractors spend on marketing percent of revenue" and recorded their ranges, citations and worked examples. We checked the SBA addresses the 7 to 8 percent figure is attributed to (both return 404) and the Internet Archive availability API (no snapshot at the paths we tried). Survey figures are from Gartner's May 11, 2026 release as reported by Marketing Dive and from The CMO Survey's 2026 Highlights and Insights report. Bid data is from the Google Ads API (v23) Keyword Planner historical metrics for 54 trade keywords, September 2025 to August 2026, Google Search only, English, geo-targeted to Canada and to the United States, in Canadian dollars. Low and high top-of-page bids are the 20th and 80th percentile of what advertisers paid; the per-trade figure is the median across that trade's keywords. Cost per lead is bid divided by LocaliQ's 2026 home-improvement conversion rate of 8.05 percent. The worked budgets use example tickets and close rates, labelled as such.

Frequently Asked Questions