Ads Manager can show you dozens of columns, and most of them won't help you decide anything. The Meta ads metrics that matter for a small business fit on one screen: what each sale costs, how much you sell for every dollar of ads, and a handful of signals that tell you why those two numbers are good or bad. This guide explains each one in plain words, shows how they connect, and gives you a simple funnel to find exactly where you're losing sales. You'll also see which popular numbers you can safely ignore.
In short: Start with money: cost per sale against your margin, and sales value per $1 of ads. Then use the signal metrics (CPM, link click-through rate, landing page views, add to cart and checkout rates) to find where the problem is. Watch frequency for fatigue. Compare Meta's reported purchases with your real orders. Ignore likes, post engagement and "all clicks".
Three layers of metrics
Think of your numbers in three layers. The first tells you if you're making money; the second tells you why; the third warns you before things break.
Layer 1: money metrics
| Metric | What it tells you | How to read it |
|---|---|---|
| Amount spent | What you paid Meta in the period | The starting point for everything else |
| Purchases (results) | Sales Meta attributes to your ads | Compare with your store's real orders |
| Cost per sale (CPA) | Spend ÷ purchases | Must stay below your margin per sale |
| Purchase value | Total value of attributed sales | Use with spend to get return |
| Sold per $1 of ads (ROAS) | Purchase value ÷ spend | Useful, but only next to your margin |
Cost per sale is the number to anchor on. Work out your break-even cost per sale first: the most a sale can cost before you lose money. If you sell a $50 item and keep $20 after product cost, shipping and fees, any cost per sale above $20 loses money, however good the other numbers look. Our guide to break-even cost per sale walks through the math.
"Sold per $1 of ads" (return on ad spend, or ROAS) is popular, but on its own it hides your costs. A return of $3 per $1 is great for a store keeping 50% of each sale and a loss for one keeping 25%.
Layer 2: signal metrics
These tell you where in the journey people drop off:
| Metric | What it measures | What a weak number usually points to |
|---|---|---|
| CPM (cost per 1,000 impressions) | What it costs to show your ad | Competitive season, narrow audience, or low ad quality |
| Link click-through rate (CTR) | % of people who saw the ad and clicked the link | The ad (image, video, first line) doesn't grab attention |
| Cost per link click | Spend ÷ link clicks | A mix of CPM and CTR |
| Landing page views | Clicks that actually loaded your page | A slow or broken page, or accidental clicks |
| Add to cart rate | Add to carts ÷ landing page views | Price, product page, photos, trust |
| Checkout completion | Purchases ÷ checkouts started | Shipping costs, payment options, surprises at checkout |
Use link click-through rate, not the default "CTR (all)", which counts any click on the ad, including clicks to see more text or the profile.
Layer 3: health metrics
| Metric | What it tells you | When to act |
|---|---|---|
| Frequency | Average times each person saw your ad | Rising frequency with falling CTR means fatigue |
| Reach | How many different people saw it | Flat reach while spend grows = saturating the audience |
| Spend pace | Whether the budget is being spent | Budget not spending = review, audience or billing problem |
| Delivery status | Learning, active, limited, rejected | Rejected or stuck in review needs action |
Our guide to ad fatigue and frequency covers what to do when frequency climbs.
Use the funnel to find where you lose sales
A single number rarely tells you what's wrong. Follow a sale from impression to purchase and look for the step that breaks. A hypothetical week for a store selling a $45 product:
| Step | Number | Rate |
|---|---|---|
| Impressions | 40,000 | |
| Link clicks | 480 | 1.2% link CTR |
| Landing page views | 400 | 83% of clicks loaded the page |
| Add to cart | 36 | 9% of page views |
| Checkouts started | 20 | 56% of add to carts |
| Purchases | 6 | 30% of checkouts |
Spend was $300, so cost per sale is $50. The ad gets clicks and the page loads. People add to cart. But only 30% of those who start checkout finish. That's a checkout problem, not an ad problem: maybe shipping costs appear late, or there are too few payment options. Changing the ad here would waste time.
Compare with a different week where link CTR is 0.4% and everything after it looks healthy. That's an ad problem: new image, new first line, new video.
Rough diagnosis guide:
- High CPM, normal CTR: competition or a narrow audience. Broaden the audience or accept seasonal costs.
- Low link CTR: creative and copy. See Meta ad copy examples.
- Many clicks, few landing page views: page speed or accidental clicks from some placements.
- Page views but few add to carts: product page, price, photos, reviews.
- Add to carts but few purchases: shipping, checkout, payment, trust.
Meta's numbers versus your store's numbers
Meta reports purchases it can attribute to your ads within an attribution setting, typically 7 days after someone clicks and 1 day after they view an ad. Meta's help article About Facebook's attribution system explains the basics. Your store counts every order. The two will never match exactly, and that's normal. Reasons include:
- Someone saw the ad, then bought days later through Google, so both platforms claim the sale.
- Someone bought without consent to tracking, so Meta never saw the purchase.
- The pixel fired twice, so Meta counted one order as two.
- Orders that came from email or word of mouth happened to be within the attribution window.
What to do:
- Compare weekly, not daily: Meta's purchases versus your store's orders.
- If Meta reports far more sales than your store, suspect duplicate events. If it reports far fewer, suspect missing tracking. Our guide to a Meta pixel that isn't tracking sales helps you check.
- Track one "blended" number too: total store revenue ÷ total ad spend. If it holds up while you spend more, your ads are working.
How to set up columns in Ads Manager
You don't need a fancy dashboard. Set up one saved column view in Ads Manager:
- Open Ads Manager and click Columns, then Customize columns.
- Add, in this order: Delivery, Amount spent, Results, Cost per result, Purchases, Purchase value (or Purchase ROAS), CPM, Link clicks, CTR (link click-through rate), Landing page views, Adds to cart, Checkouts initiated, Frequency, Reach.
- Tick Save as preset and give it a name.
- Set the date range to the last 7 days for weekly reviews, and compare with the previous period.
Check that the results column shows what your campaign optimizes for: purchases for a sales campaign, leads for a lead campaign.
Metrics you can safely ignore
- Likes, reactions and post engagement on sales ads. Nice to see, not money.
- CTR (all) and "clicks (all)": they include clicks that never reach your site.
- Reach on its own: reaching many people who don't buy is just expensive awareness.
- Relevance-style rankings in isolation: quality, engagement and conversion rankings compare you with other advertisers' ads and only show up after enough impressions. Useful as hints, not as goals.
- Daily swings: one great or awful day means little. Look at 7-day windows.
What if…? Reading confusing numbers
What if CTR is great but there are no sales? The ad works; the page or offer doesn't. Check landing page views, add to cart rate and checkout. Also make sure the ad doesn't promise something the page doesn't show.
What if the cost per sale jumped overnight? Check for a recent edit, a holiday or sales season, a rejected ad, or a tracking problem before changing anything. Our guide to a cost per sale spike has a step-by-step checklist.
What if Meta's return looks great but your bank account doesn't? Compare reported purchases with real orders, check for duplicate purchase events, and calculate profit with your real margin, not revenue.
What if a campaign shows zero results but spent money? Check that the results column matches the objective, that the pixel is recording the event, and that the campaign has had enough time. If it spent two to three times your break-even cost per sale with no sales, pause it.
How Meta Ads Copilot Max reads your metrics for you
Meta Ads Copilot Max reads your campaigns, ad sets and ads from Ads Manager (or a CSV export) and shows the numbers that matter: spent, results, cost per result, impressions, clicks and CTR, for the period you choose. Results follow each campaign's objective, so purchases, leads and messaging conversations are each counted the right way.
It then gives the account a score from 0 to 100 and plain-language alerts: spending with no results, each sale costing more than your margin, low click-through rate, high frequency, a budget that isn't spending, ads in review or rejected, and missing purchase tracking. Each analysis is saved with its history. Tick an alert and Claude in Chrome (required) makes the fix in your browser, with nothing published or spent without your approval.
You can try Meta Ads Copilot Max free for 7 days. Start your free 7-day trial and we'll email you the key and the download link.
Frequently asked questions
What is the most important Meta ads metric?
For a store, cost per sale compared with your margin per sale. It tells you directly whether each sale from ads makes or loses money. Every other metric helps explain why that number is where it is.
What is a good CTR for Meta ads?
It varies a lot by product, audience and placement, so be careful with generic benchmarks. Compare your ads with each other and with your own history. As a rough working rule, a link CTR below about 1% after a few thousand impressions is worth a creative review.
What does CPM mean in Meta ads?
CPM is the cost of showing your ad 1,000 times. It rises in competitive seasons like Black Friday and with narrow audiences. A high CPM isn't a problem on its own if your cost per sale stays below your margin.
Why don't Meta's purchases match my store's orders?
Meta only counts purchases it can attribute to an ad within its attribution setting, and some purchases can't be tracked at all. Duplicate pixel events can also inflate Meta's count. Compare the two weekly and investigate big gaps.
How often should I check my Meta ads metrics?
A quick daily look for problems (rejected ads, spending with no results) and a proper weekly review on 7-day numbers is enough for most small businesses. Making decisions on daily numbers usually leads to over-reacting.
