TL;DR
- Meta's ad auction ranks ads by total value: the bid, Meta's estimated action rate for that person, and ad quality, so a larger bid alone does not win.
- Meta says an ad set usually exits the learning phase after about 50 results in the week after its last significant edit.
- Meta's standard attribution can credit a purchase up to 7 days after a click or 1 day after a view, so its counts often differ from Google Analytics and shop orders.
- Break-even ROAS equals 1 divided by gross margin: an illustrative product with a 40% margin needs a ROAS of 2.5 to cover the first order.
Facebook ads are worth it when you have an offer people will buy without searching for it first, a margin that can absorb the cost of reaching them, and enough budget and tracking to let Meta's delivery system learn. They are rarely worth it for a product with thin margins, a tiny audience, a weak landing page or no working conversion tracking.
That is the short answer to "are Facebook ads worth it". In practice the question depends more on your business than on the platform. This guide explains how Meta's ad system actually decides who sees your ads, what a fair test looks like, how to judge results against your own numbers, why campaigns stall, and how Facebook compares with Google Ads for the same budget. Any numbers below are illustrative examples to show the maths, not benchmarks.
How Facebook ads work, in brief
Every time someone opens Facebook or Instagram and there is space for an ad, Meta runs an auction. It does not simply pick the highest bidder. According to Meta's own description, the winning ad is the one with the highest total value, which combines three things:
- Your bid, or the bid strategy you set, such as highest volume or a cost per result goal.
- Estimated action rates, which is Meta's prediction of how likely this person is to take the action you are optimising for, such as a purchase or a lead.
- Ad quality, judged from feedback such as people hiding the ad and from signals of low-quality attributes like engagement bait or withheld information.
This matters for the "worth it" question because it means money alone does not buy results. A relevant ad with good creative, aimed at an action people actually take, can win auctions at a lower cost than a bigger bid on a poor ad. It also means Meta needs data on the action you choose. If it cannot see purchases or leads, it cannot predict them.
Objectives decide what Meta optimises for
Meta Ads Manager groups campaigns into six objectives: Awareness, Traffic, Engagement, Leads, App promotion and Sales. The objective tells the system who to find. A Traffic campaign looks for people likely to click. A Sales campaign optimised for purchases looks for people likely to buy, which is a narrower group.
Running the wrong objective can make Facebook ads look as if they are not worth it. A Traffic campaign can bring cheap clicks that never buy, because buying was never what it was told to find. If the business goal is sales or leads, the campaign objective and the conversion event should match it.
When Facebook ads make sense, and when they don't
Facebook and Instagram are discovery platforms. People are scrolling, not searching. That makes them strong for some businesses and weak for others.
They tend to make sense when:
- The product is easy to understand from an image or a short video, such as fashion, food, beauty, home goods, courses or events.
- The audience is broad enough for the system to find buyers. A national consumer product has room; a service for a few hundred companies usually does not.
- The margin per sale, or the lifetime value of a customer, is high enough to pay for the clicks and views it takes to win one.
- People buy on impulse or after a short consideration, so an ad can move them from seeing to buying in days, not months.
They are harder to make worth it when:
- Demand is urgent and people search for it, such as emergency repairs, where search ads catch intent that social ads have to create.
- The audience is very narrow or very senior, and the platform's targeting cannot reach it efficiently.
- The product is in a restricted category under Meta's advertising standards, such as some financial, health or alcohol products, where approval and targeting have limits.
- The landing page or checkout is slow, unclear or broken. No campaign fixes a page that does not convert.
How to judge results: CPA and ROAS against your margin
Judge Facebook ads by what a customer costs compared with what that customer is worth to you. Likes, reach and even click-through rate are diagnostic numbers, not results.
Two measures do most of the work:
- CPA (cost per acquisition): ad spend divided by the number of purchases or leads.
- ROAS (return on ad spend): revenue attributed to ads divided by ad spend.
Neither means anything until you compare it with your own margin. Here is an illustrative example, not a benchmark. Say you sell a product for 1,500 baht and your gross margin after product cost and delivery is 40%, or 600 baht.
- Your break-even CPA is 600 baht. If a sale costs more than 600 baht in ads, each first purchase loses money.
- Your break-even ROAS is 1 divided by the margin, so 1 / 0.40 = 2.5. Below a ROAS of 2.5, ads lose money on the first order.
Repeat purchases change the picture. If a typical customer buys three times in a year, a first-order CPA above 600 baht may still be worth paying, as long as you can see the repeat purchases in your own data. That is a business decision you can only make with your numbers, not with industry averages.
Attribution windows: what Meta counts as a result
Meta reports a conversion when it happens within an attribution window after someone clicks or views your ad. With Meta's standard attribution model, the setting can count a website purchase that happens within 1 or 7 days of a link click, within 1 day of an ad impression (view-through), or within 1 day of a non-link engagement such as a 5-second video play.
That has two consequences. First, Meta's numbers will often differ from Google Analytics, which uses its own attribution and does not see views. Second, view-through conversions can make results look better than they are, because some of those people would have bought anyway. A sensible habit is to compare Meta's reported purchases with the orders in your shop system over the same period, and to check how results look with a 1-day click setting using Ads Manager's compare attribution settings feature. If the business only looks healthy with view-through conversions included, be careful.
What a realistic test budget and test length look like
A fair test gives Meta enough conversions to learn and runs long enough to cover normal weekly swings. Meta's guidance says an ad set usually leaves the learning phase after about 50 results in the week after its last significant edit.
You can use that to size a test. Again, this is illustrative maths, not a benchmark. If you expect a purchase to cost about 300 baht, then 50 purchases a week is around 15,000 baht a week, or a little over 2,100 baht a day, for one ad set. If that is more than you can spend, you have three honest options:
- Optimise for an event that happens more often, such as add to cart or initiate checkout, and accept that it is a weaker signal of real sales.
- Use fewer ad sets and fewer audiences, so the budget is not split into pieces too small to learn.
- Accept that the account will stay in "Learning limited" status and that results will be less stable, and judge it over a longer period.
As for length, plan on at least two to four weeks before drawing conclusions, and longer if the purchase cycle is long. Judging a campaign after three days, or changing it every day, usually tells you more about the learning phase than about whether Facebook ads work for your business.
Why Facebook ad results stall
When results get worse after a good start, check a few mechanical problems before deciding the platform has "stopped working".
The learning phase keeps resetting
Significant edits put an ad set back into learning. Meta lists changes such as editing targeting, creative, the optimisation event or bid strategy, adding a new ad, pausing for seven days or more, and large budget changes. Teams that edit every day keep the system in permanent learning, where costs are higher and less stable.
Creative fatigue
Watch the frequency metric, which shows how many times on average each person has seen your ads. As the same people see the same ad again and again, click-through rate tends to fall and cost per result tends to rise. The usual fix is new creative, such as a different angle or opening, refreshed on a regular schedule.
A weak landing page
If click-through rate is healthy but conversion rate on the site is low, the problem is after the click. Slow mobile pages, unclear pricing, a long form or a checkout that asks for an account first will waste spend that the ad did its job to earn.
Broken or partial tracking
Since Apple's App Tracking Transparency arrived with iOS 14.5 in 2021, the Meta Pixel in the browser sees fewer conversions on its own. Meta's Conversions API sends events from your server and fills part of that gap. If the Pixel fires twice, fires on the wrong page, or misses purchases altogether, the delivery system is optimising on bad data. Check Events Manager for the event match quality score and for duplicate or missing events before deciding the platform failed.
Facebook ads vs Google Ads for the same budget
The difference is intent. Google Search ads show to people who are already looking for what you sell. Facebook ads show to people who match a profile of likely buyers but were not looking at that moment. The comparison below shows how that plays out.
| Factor | Facebook and Instagram ads | Google Search ads |
|---|---|---|
| How demand is reached | Creates demand among people not yet searching | Captures demand from people already searching |
| Main lever for results | Creative and offer | Keywords, bids and landing page match |
| What the system learns from | Conversion events from Pixel and Conversions API | Conversion tracking plus search query intent |
| Best fit | Visual, impulse or new-to-market products | Urgent needs and products people already know to search for |
Many businesses end up using both: search to catch people ready to buy now, and social to build the audience that will search later. If you are also weighing search, the Google Ads service page covers how that side is managed. With a small budget, start where your buyers already are and add the second channel once the first pays for itself.
What this means for Thai businesses
For Thai businesses, the harder questions are often about the sales process after the ad.
Click-to-message ads open a conversation in Messenger, Instagram Direct or WhatsApp. They suit businesses that sell through chat, but a chat only counts once it turns into an order. If a Thai business optimises for conversations, it needs a way to record which chats became orders, or the ads will look either better or worse than they are. The same applies to shops that close sales on LINE: Meta cannot see those orders unless you send them back through the Conversions API.
Language also matters. Write Thai creative for Thai buyers instead of translating English copy line by line, and keep the ad, landing page and chat replies in the same voice. A bilingual business may need separate ad sets for Thai and English audiences so each one can be read on its own.
If your team does not have time to manage tests, creative refreshes and tracking, that is a reasonable sign it is time for outside help. A specialist Facebook ads service can set up the account structure and tracking, and Instagram-led brands may fit better with dedicated Instagram ads management. Businesses running several social platforms together can look at paid social campaigns that plan Meta, TikTok and other channels as one budget.
FAQ: are Facebook ads worth it?
Are Facebook ads worth it for small businesses?
They can be, if the business has a clear offer, a working website or chat sales process, and enough budget to give one or two ad sets around 50 conversions a week. With less budget, keep the account simple and judge results over a longer period.
How long should I test Facebook ads before deciding?
Plan for at least two to four weeks without major edits. Shorter tests mostly measure the learning phase, and frequent changes reset it.
Why are my Facebook ads getting clicks but no sales?
The usual causes are the wrong campaign objective, a landing page that does not convert, or tracking that does not record purchases. Check that the campaign is optimised for purchases or leads, then test the page and the Pixel before changing the ads.
When is it time to hire a Facebook ads agency?
It is usually time when spend is high enough that small improvements matter, when tracking or creative is holding results back, or when nobody in-house has hours each week to manage the account. An agency should be able to explain its testing plan and how it will measure results against your margin.
Facebook ads are worth it when the numbers behind them work for your business, and you can usually find out within a month or two of a well-run test. If you would like help setting up that test, Relevant Audience can review your offer, tracking and account structure and suggest a plan that fits your budget.
