A sane Facebook & Instagram strategy
Google Ads answers demand that already exists — someone types what they want and you show up. Meta is the opposite. Nobody opens Instagram intending to buy from you. That single difference is why so much small-budget Meta spend evaporates: it's being run as though it were search, and it isn't. Interrupting someone successfully is a different craft, and it has its own discipline.
On Meta, the creative is the targeting
This is the part that has genuinely changed. Meta's machine learning is now better at finding the right person than you are at describing them — and a decade of obsessive interest-stacking has quietly stopped being the lever it once was. Give the system a sensible audience and a clear conversion signal, and it will go looking. What it can't do is make a boring ad interesting. Your creative is now the strongest instruction you give the algorithm: the people who stop for it are the people it will find more of.
Which means the hours you'd have spent tuning audiences are better spent making more, and more varied, things to say.
Test concepts, not colours
The classic small-budget mistake is running five near-identical ads and waiting for a winner. They'll all perform about the same, because they're all the same idea. Test ideas that could each be right for a different reason: the problem stated bluntly, the finished result, the price objection met head-on, the thirty seconds of a real person explaining what they do. One of those will outperform the others by a margin that no headline tweak would ever produce.
Cold, warm, and the discipline of the funnel
Someone seeing you for the first time and someone who read your pricing page last week are not the same person, and shouldn't get the same ad. Keep it simple: something that earns attention and explains who you are for people who've never heard of you, and something direct — an offer, a booking, a reason to act now — for people who've already shown interest. That's the whole funnel most small businesses need.
Retargeting is where budgets go to die quietly
Retargeting looks brilliant in the reporting, because it takes credit for people who were coming back anyway. It's still worth doing — but cap the frequency, exclude anyone who's already bought or enquired, and be honest that its numbers flatter it. Following someone around the internet for six weeks isn't persuasion. It's just annoying.
Judge it on the business, not the dashboard
Since iOS privacy changes, Meta's in-platform attribution is an estimate with an optimistic streak. Don't fight it — triangulate. Watch what happens to total enquiries and total revenue when spend goes up and when it goes down. Ask new customers how they found you. A slightly blurry read on reality beats a confident number that isn't true.
Where to start with a small budget
- Get conversion tracking working properly before you spend — everything downstream depends on it.
- Make four genuinely different creative concepts, not four variations of one.
- Run one broad prospecting campaign. Resist the urge to slice it into fragments.
- Add one modest retargeting campaign, with exclusions and a frequency cap.
- Kill what's clearly losing, feed what's clearly working, and let the rest run long enough to mean something.
Meta rewards patience and good creative, and punishes fiddling. Most small budgets aren't too small — they're just spread across too many things, changed too often, and judged on the wrong numbers.