Facebook Retargeting for Small Businesses: How to Win Back Website Visitors Without Wasting Money
Picture a Hamilton landscaper who gets 400 visits to her website in a month. Maybe 12 people fill in the quote form. The other 388 look at a few photos, think "I'll sort that out later," and close the tab.
Most of them won't come back by themselves. Retargeting is how you give them a gentle nudge. You show your ads on Facebook and Instagram only to people who already visited your site. They know who you are, so you aren't starting from zero.
It works well. It also goes wrong in a few boring ways, and small businesses hit them more than big ones.
Why it's worth doing
Cold audiences are expensive. You pay to introduce yourself to strangers who didn't ask. People who've already been on your site are a short step from enquiring.
One agency that audits Meta accounts puts retargeting returns at 5 to 10 times what you spend, compared with around 1.5 times for ads aimed at new people (that figure comes from food and drink brands, so treat the exact numbers loosely). The direction is the useful part: warm people cost less to win than cold ones.
The catch is that retargeting can only finish a job that something else started. If nobody visits your site, there's nobody to retarget. So it sits on top of your other ads, not instead of them. It's the same logic as the setup we walk through in our full guide to Google and Meta ads: one part finds people, another part brings back the ones who hesitated.
The small audience problem
Here's what big-business advice skips. If your site gets 400 visits a month, your retargeting audience is tiny. Some of those visits are you, your staff and people who bounced in two seconds.
Tiny audiences cause two problems. Meta struggles to learn anything from them. And a few hundred people will see the same ad again and again until they're sick of it.
A few ways to handle this:
Lump the audience together. Don't build a separate campaign for every page. One audience of "everyone who visited in the last 30 days" is plenty to start. The agency we looked at suggests keeping any group under about 5,000 people in one campaign.
Use a longer window than you think. If you only get a few hundred visits a month, a 7-day window leaves almost nobody to show ads to. Try 30 days, then 60 or 90 if the numbers are still thin. Past about 180 days the results tend to fade, because people forget they ever looked.
Add people who engaged with you on Facebook and Instagram. Meta lets you include people who watched your videos, visited your profile or messaged you. That grows the pool without needing more website traffic.
Keep your ad budget small, and cap how often people see it
Retargeting should be the smaller part of your Meta spend. A common rule is around 15 to 25 percent, with the rest going to ads that find new people. A tradie spending $600 a month on Meta might put $100 to $150 into retargeting.
Then check frequency, which is how many times the average person has seen your ad. If it creeps past five or six in a couple of weeks, you're mostly annoying people. Either widen the audience, refresh the ad, or drop the budget.
Don't show the same ad to everyone
This is where a lot of small accounts fall flat. The ad that worked on strangers ("Need a landscaper? Get a free quote") is the wrong ad for someone who's already seen your work and still hasn't enquired.
Think about why that person left. Usually one of a few things: they weren't sure of the price, they weren't sure you'd be any good, or it just wasn't urgent. So answer that worry directly.
For the price worry, show a real example job with a real dollar range.
For the trust worry, show a short customer story or before-and-after photos.
For the not-urgent worry, give them a reason to act this month, like a limited booking slot or a bonus that ends on a set date.
Keep a few different ads running and swap them every two to three weeks. The agency we mentioned found that giving Meta plenty of creative variety beat splitting the same budget into lots of tiny ad sets.
Leave out the people you don't want to pay for
Before you launch, exclude three groups:
People who already became customers or sent an enquiry. Showing a "get a quote" ad to someone who already got one is a quick way to look careless.
Your own team. If you click through your own ads, you're paying for it.
Shorter-window audiences from longer ones. If you run a 30-day and a 90-day audience, take the 30-day people out of the 90-day group, or they'll be in both and see double.
Set up in about an hour
Check that the Meta pixel (a bit of code that tells Meta who visited your site) is installed and firing on your key pages. Build one audience of website visitors from the last 30 days, and take out anyone who reached your "thank you" page after enquiring. Make one campaign with a small daily budget, say $5 to $10, and two or three ads that each answer a different worry. Look at it after two weeks, not two days.
Judge it by enquiries and jobs, not clicks. If you're unsure whether your tracking even counts enquiries properly, fix that first, because retargeting is only as good as the data behind it.
