Why your Facebook
ads aren't working.
Paid media · 9 October 2026 · 5 minute read
Search ads reach people already looking for what you sell. Meta ads interrupt people who were not looking for anything. That one difference changes what counts as a failure and what fixes it.
When Meta ads do not work it is almost always the creative or the offer, not the targeting. Targeting is where people spend their time because it is the part with buttons.
The difference that explains most of it
On Google, someone types what they want and you pay to appear. The intent already exists and your job is to be the obvious answer.
On Meta, nobody asked. Your ad appears between a friend's holiday photos and a video of a dog. Your job is first to interrupt, then to make a case, then to ask. That is three jobs, and most underperforming Meta ads fail at the first one and never get a chance at the other two.
This is why advice imported from search advertising tends not to help here. Keyword logic, match types and bid adjustments are answers to a problem Meta does not have.
It is usually the creative
The platform's own automation has become good at finding the right people, provided you give it something worth showing them. Which means the lever that actually moves performance is the thing in the ad, not the settings around it.
Three questions about your creative, in the order they matter.
- Does it stop the scroll in the first second? Not the first three seconds. The first one. A static image of your logo and a discount does not.
- Does it look like the platform or like an advertisement? Ads that match the visual grammar of the feed consistently outperform ads that announce themselves as ads. This is not about being deceptive, it is about not being skipped before you are read.
- Does it say something specific? "Quality service, great prices" is interchangeable with every competitor. A specific claim, a real situation, a number, an actual thing you do, is what makes someone pause.
If you have run the same three images for six months, you do not have an ad account problem. You have a content problem.
Creative fatigue is real and it is fast
A search ad can run for a year because every person who sees it is seeing it for the first time, at the moment they searched. A Meta ad is shown repeatedly to the same relatively small local audience. After a few weeks the same people have seen it many times, performance decays, and the platform has to pay more to find someone who has not.
The symptom is an ad that worked, then quietly stopped, with nothing changed. The fix is new creative rather than new targeting. For a local business with a small audience, assume you need fresh variations regularly, not once a quarter.
Stop narrowing the audience
Detailed targeting felt like the whole craft a few years ago. It is now mostly a way to make the system's job harder.
A local service business usually does best with a broad audience, a location radius, a sensible age range if it genuinely matters, and nothing else. Stacking six interests on top of a small local population leaves the system too little room to find the people who convert.
The exception is retargeting, which is narrow by definition and still worth running: people who visited the website, people who engaged with the page, your customer list. These audiences are small and they convert at a completely different rate to cold traffic, so keep them in separate campaigns or the reporting will mislead you.
The offer problem
Because Meta interrupts rather than answers, the thing you ask for has to be proportionate to how little the person was thinking about you ten seconds ago.
"Book a consultation" is a large ask of a stranger. Something smaller and more concrete, an actual first step with an obvious value to the person taking it, converts a cold audience far better. This is why free assessments, guides and first-visit offers exist, and why a generic "contact us" rarely works on cold traffic.
If your Meta ads produce clicks and no enquiries, test the offer before you test the audience.
Check the tracking, again
Meta reporting depends on the pixel, the conversions API, or both, and on browser and operating system privacy settings that have made the picture less complete than it used to be. Two consequences worth holding in mind.
- Meta usually under-reports. Conversions it cannot attribute do not disappear from your business, only from the dashboard. Compare against your own enquiry count.
- Meta and Google will disagree. Both will claim the same customer. Attribution windows differ, models differ. Pick one source for decisions and treat the rest as directional.
Before concluding that Meta does not work for your business, confirm that the enquiries it is producing are being counted at all. A business that takes bookings by phone and judges Meta on form fills will reliably conclude the wrong thing.
Give it enough budget to learn
Meta's delivery system needs a certain volume of the event you are optimising for before it performs predictably. Below that, it is effectively guessing and your results swing week to week for no reason you can identify.
Running four campaigns on a small total budget guarantees that none of them reaches that point. One campaign, properly funded, will teach you more than four that are each starved.
A checking order
- Confirm enquiries from Meta are actually being recorded, including phone calls.
- Look at the creative. Is it specific, does it stop the scroll, how old is it.
- Check how long the current creative has been running and how often the audience has seen it.
- Widen the targeting rather than narrowing it, and separate retargeting from cold.
- Make the offer smaller and more concrete.
- Consolidate budget into fewer campaigns.
- Judge the result against your own enquiry count, not only the dashboard.
Common questions
How much do I need to spend on Facebook ads to see results?
Enough that the campaign reaches a steady volume of the action you are optimising for, which depends on your cost per result rather than a fixed figure. The more useful discipline is to put the whole budget behind one campaign rather than splitting it across several that each learn nothing.
Why did my Facebook ads stop working?
Most often creative fatigue. Meta shows your ad repeatedly to the same audience, so after a few weeks the people most likely to respond already have, and performance decays with nothing having changed. New creative usually fixes it. New targeting usually does not.
Should I boost posts or run proper campaigns?
Boosting is convenient and gives you almost no control over objective, placement or measurement, so it tends to buy engagement rather than customers. If the goal is enquiries, run it as a campaign with the objective set to the thing you actually want.
Do Facebook ads still work for small businesses?
Yes, for the right kind of offer. They work well where the purchase is visual, relatively quick to decide on, and where demand can be created rather than only captured. They work less well for services people only buy at the moment they search, which is where Google is stronger.
Where to start
Not sure which of these is costing you the most?
I'll look at your website, your social presence and how you show up in local search, and tell you plainly which one is the problem. Written findings in five business days. All I need is your website address.