You have a number. Maybe it is five hundred a month, maybe it is two thousand, but it is a real number, and it came out of a conversation with yourself in the truck about how much you can lose for a few months without it hurting. Now the question is where it goes.
That is the part nobody answers plainly. Every channel has a salesperson, and every salesperson thinks your whole budget belongs to them. The Google rep wants it in ads. Angi wants it in leads. The guy who built your cousin's website wants it in a website. None of them are lying, exactly. They are just answering the wrong question. The right one is not "which channel is best," it is "in what order, and how much to each, for a business your size."
Here is a way to split a contractor advertising budget that holds up whether the number is small or large. It is built on one rule: spend on the things you own before you rent anything.
Start with what you own, because it makes everything else cheaper
Two things belong to you outright: your Google Business Profile and your website. Everything else — search ads, marketplace leads, boosted posts — is rented. The rent stops the day you stop paying.
The reason owned assets come first is not sentiment. It is that every rented dollar lands on them. Someone clicks your ad and reaches your site. Someone sees you on the map and reads your reviews. If the site is slow or has no phone number above the fold, or the profile has eleven reviews and the last one is from 2023, the rented traffic leaks straight out the bottom. You are paying full price to send people somewhere that does not convert them.
So the first slice of any budget, and the first month of any plan, goes here:
- Google Business Profile. Free to set up, and the single highest-return listing a local contractor has. Fill in every field, add photos of real jobs, and start asking for reviews at the moment the customer is happiest. The reviews post covers exactly when and how to ask.
- A website that gets you called. Not a brochure. A site with the phone number visible without scrolling, a quote form that works on a phone, your service area spelled out, and your best before-and-afters. This is a one-time cost, not a monthly one, which is the whole point.
If the number is small, this is where most of it goes for the first month or two, and that is correct. You are not "not advertising." You are fixing the bucket before you turn on the tap.
Then reviews, because they are the cheapest lead source you have
Reviews are marketing, and they cost you nothing but discipline. A contractor with sixty recent reviews at 4.8 beats a contractor with twelve at 5.0, in the map, in the ads, and in the phone call — because the homeowner already trusts you before you pick up.
Put a small standing amount here, mostly in time rather than money: a text template, a link that takes one tap, and the habit of sending it the day the job wraps. If you want to spend actual dollars, a review-request tool runs a few dollars a month and is worth it only if it makes you consistent. Consistency is the product; the tool is just a reminder.
This slice never goes away. It is the one part of the budget that keeps paying after you stop.
Now the rented channels, in the order that makes sense
Once the owned pieces are working, the remaining money goes to channels that bring strangers. There are two big ones for most trades, and they behave very differently.
Google Ads puts you above the map for the exact search a homeowner types when they need you today. The Google Ads post has the current benchmark costs and the arithmetic, so this post will not repeat them. The short version: it is an auction, clicks in the home-improvement category are expensive, and it works best for big-ticket work where one job covers a month of spend. It also has a floor. Spend too little and you never buy enough clicks to learn anything, then you conclude it does not work.
Marketplaces like Angi and Thumbtack send you leads without you building anything, which is why they are tempting on day one. The Angi and Thumbtack post walks through what they charge, how shared leads change the math, and where they genuinely help. The short version: they are a fast on-ramp when the phone is silent, and an expensive habit when it is not, because you are renting customers you could have owned.
Which gets the bigger share depends on one question: what is your average ticket?
- Big tickets — roofs, full remodels, HVAC replacements, anything where one job clears several thousand dollars. Google Ads gets the larger share. A lead that costs a hundred dollars is cheap when the job is eleven thousand. Marketplaces can fill gaps but should not be the main line.
- Small tickets — service calls, repairs, anything under a few hundred dollars. Paid search rarely survives the math. Here the budget leans harder on the owned assets and reviews, with marketplaces used sparingly to fill empty days rather than as a strategy.
Either way, do not run both rented channels at full tilt in the same month with a small budget. Pick one, give it a real three-month test at a level where you can learn something, then decide.
A worked split for three budgets
These are not rules, they are starting points to argue with. The percentages assume the website is already built; treat the build as a separate one-time line before any of this.
$500 a month. Almost all of it is owned and reviews. Spend on the website fixes you have been putting off, a review tool if you need the nudge, and a little on your Google Business Profile photos. Do not start Google Ads here; you cannot fund it at a level that teaches you anything. If you need work right now, a tight marketplace budget can fill the gap for a month or two while the owned pieces catch up.
$1,000 a month. Roughly a fifth stays on owned and reviews as a standing maintenance line. The rest goes to one rented channel, chosen by ticket size. For a big-ticket trade that is Google Ads at a level where you actually buy a meaningful number of clicks. For a small-ticket trade it is a marketplace, watched closely, with the money pulled the month it stops paying.
$2,000 and up. Now you can run both rented channels and compare them, still with a standing slice on owned and reviews. Track cost per booked job, not cost per lead, for each channel. After a quarter, move money toward whichever one books cheaper. That reallocation, done every quarter, is the entire strategy.
Track one number per channel, or you are guessing
The number is cost per booked job. Not clicks, not impressions, not leads. A lead is someone who filled in a form; a job is someone who paid you.
To get it, you need to know where each job came from. That means a different phone number or a "how did you hear about us" question on every rented channel, and the discipline to write it down. It is unglamorous, and it is the only thing that makes the quarterly reallocation possible. Without it you are moving money based on which rep called you last.
When this split does not apply
If you are booked out six weeks, you do not need rented leads at all. Put the whole budget into owned assets and reviews, and let the waiting list grow. Spending on ads when you cannot take the work just trains customers to call someone else.
If you are brand new with no reviews and no site, the marketplace slice is larger than this post suggests, for a short while. You need a few real customers before the owned assets have anything to show. Just set an end date for it.
And if your trade is so specialised that nobody searches for it, paid search is off the table entirely and referrals carry the load. The split still applies, it just skips a channel.
Do this one thing this week
Write your number at the top of a page. Under it, write your average ticket. Then write the four lines — owned, reviews, Google Ads, marketplaces — and put a dollar figure next to each one for next month. Fixed number, real allocation, one channel test at a time. That page is more marketing strategy than most contractors ever write down, and it took you five minutes in the truck.