Meta ads are effective for a small business when three conditions hold: your margin can pay for finding customers who were not looking for you, you can generate enough conversions each week for Meta's delivery to learn, and you can supply fresh creative every few weeks. When all three hold, Meta is one of the best discovery channels a small business can buy. When one fails, the ads usually lose money however well they are set up.
That is a more useful answer than "it depends", because each condition can be checked before you spend anything.
The checklist
Answer each honestly. Two or three yes answers make Meta worth testing; a no is something to fix first or a reason to choose another channel.
| Condition | Yes looks like | No looks like | If no, do this |
|---|---|---|---|
| Margin pays for discovery | Gross margin per order comfortably above a realistic cost per customer | Margin of a few euros per sale | Raise order value with bundles, or advertise a higher-margin product |
| Enough conversion volume | A steady stream of purchases or leads every week from ads | A handful of sales per month | Optimise for a higher-funnel event, or use search ads first |
| Creative supply | Several new ads every few weeks | One ad, made once | Build a footage library and a refresh routine before scaling |
| A product people discover | Visual, impulse-friendly, novel | Bought only when needed and searched for | Put most budget into search; use Meta for retargeting |
| A landing page that converts | Fast, one offer, matches the ad | Home page, slow on mobile | Fix the page before judging the ads |
Condition 1: the margin has to pay for discovery
Meta finds people who did not know you existed. That costs more per customer than catching someone already searching, and the first purchase has to cover it, or repeat purchases have to.
The arithmetic is short. Your break-even ROAS is one divided by your margin. A business with a 40 percent gross margin breaks even at a ROAS of 2.5: every euro of ad spend must bring back 2.50 in revenue before the ads make any money. At a 20 percent margin the break-even ROAS is 5, which is hard to hit consistently on cold audiences. Run your own numbers in the break-even ROAS calculator.
If repeat purchase is strong, judge the first order against a lower bar, because the customer is worth more than one sale. Be honest about whether repeat purchase is real or hoped for.
Condition 2: enough conversions for the algorithm to learn
Meta's delivery system learns from the conversion event you choose. Give it a steady run of those events every week and it gets better at finding buyers. Give it a few a month and it never leaves the learning phase, so delivery stays erratic and costs stay high.
Small businesses run into this all the time, and the fix is not more patience. Optimise for an event that happens more often:
- Add to cart or initiate checkout instead of purchase, for a shop with low order volume.
- Lead form submission instead of booked appointment, for a service business.
- Landing page views as a last resort, which teaches Meta to find people who click and stay, not buyers.
Move back to purchase optimisation once weekly volume grows. If spend itself barely moves, that is a different problem with its own causes, covered in why Meta ads are not spending.
Condition 3: creative volume to survive fatigue
This is the condition small businesses underestimate most. Small audiences see the same ad repeatedly, and ads wear out. Our 2026 fatigue benchmark reports CTR dropping 15 to 20 percent in the first two weeks, week three landing 45 to 70 percent below launch, and a 45 percent CTR drop after a fourth exposure to the same creative. It puts a typical active campaign at 8 to 20 live variations.
A small budget cannot support 20 live ads, because each needs enough impressions to read. The realistic small-business version is three or four live at a time with two new ones every week or two, worked out in our post on ad variations on a small budget. What matters is the habit: a business that makes one ad and runs it for three months will conclude Meta does not work, when the ad simply died in week three.
Where Meta ads fail small businesses
Some businesses should not lead with Meta, and saying so saves money:
- Very local businesses with a tiny radius. A dentist serving one neighbourhood has an audience too small for Meta to optimise over. Google Business Profile and search ads usually beat it.
- High-consideration purchases with long cycles. A kitchen renovation is researched for months. Meta can build awareness, but judging it on direct purchases in a month will make it look like a failure.
- Tiny margins. A product with a few euros of margin cannot pay for discovery unless order value rises.
- Products people only buy when a problem arrives. Nobody scrolls into needing a locksmith. Be present when they search instead.
For these businesses Meta still has a role, usually retargeting people who already visited the site, with a small budget.
How to test Meta properly on a small budget
A fair test, so you know whether Meta works for you rather than whether one ad did:
- Fix tracking first. One conversion event, tested by making a purchase or submitting a lead yourself.
- One campaign, one audience, three ads. Broad targeting usually works better than narrow interests at small budgets because it gives delivery more room.
- Read hook rate after a couple of days. Platform medians are 28 percent on Meta feed and Reels in our benchmark. Replace the weakest ad.
- Read cost per customer after a few weeks, against your break-even, not against a number someone quoted on a forum.
- Refresh creative every two to three weeks and give the test at least six weeks before deciding.
The full setup, including how to structure the campaign so the test is readable, is in our Meta creative testing setup.
Where Genyad fits
Condition three is the one we work on. Genyad turns footage you already own into video ad variations, each with its own script, shot selection, voiceover and captions, exported at 9:16, 4:5, 1:1 or 16:9 for Meta placements. The free plan covers 5 video ad variations plus one AI-generated video of up to 10 seconds, with no card, and paid credit packs start at €29 for 15 credits. More on how it handles Meta is on our Meta ad creative page.
What it does not do: no static image ads, no AI avatars, no predicted performance scores and no publishing into Meta on your behalf. You upload the exports to Ads Manager yourself. It also cannot fix conditions one and two; no creative tool can make a thin margin profitable.
Frequently asked questions
Do Facebook and Instagram ads work for small businesses?
Yes, when the margin covers the cost of finding new customers, the account produces enough conversions each week for delivery to learn, and new creative arrives every few weeks. If any of those is missing, fix it first or start with search ads.
How much should a small business spend on Meta ads?
Enough to get a readable result from a few ads over several weeks, which for most small businesses means a few hundred euros a month. Spreading less than that across many ads or audiences produces data too thin to act on.
Why are my Meta ads not working?
The common causes, in order: broken conversion tracking, a conversion event too rare for delivery to learn from, creative that has worn out, and a landing page that does not match the ad. Check them in that order before changing targeting.
Should a small business use Meta ads or Google ads?
If customers search for what you sell, start with Google search ads, which reach people already looking. If your product is something people discover rather than search for, Meta is usually the better first channel. Many businesses end up using both: search for intent, Meta for discovery and retargeting.