Meta Ads KPIs Explained: Every Metric That Matters
Meta ads KPIs are the metrics inside Ads Manager that tell you whether a campaign is working — split into delivery metrics (CPM, reach, frequency), engagement metrics (link CTR, landing page views), and outcome metrics (cost per result, ROAS). None of them means much alone. A cheap CPM with a weak link CTR is still a losing campaign.
Call them Facebook ads metrics or Facebook ad KPIs — same numbers. This guide covers what each is built from, where people misread it, and which Facebook ad metrics matter for your account.
Which Meta Ads KPIs Actually Matter?
Group them into three layers, top to bottom of the funnel:
Delivery — is the ad getting shown, and to how many people? Impressions, reach, frequency, CPM.
Engagement — is the ad getting clicked and the page loading? Link clicks, link CTR, CPC (link click), landing page views.
Outcome — is the result actually worth the spend? Cost per result, CPA, ROAS, website purchase ROAS.
Outcome metrics alone tell you that something is broken. Delivery and engagement tell you where.
What Do the Delivery Metrics Mean?
Impressions count every time your ad appeared on screen. Repeats count — one person seeing an ad three times is 3 impressions, not 1.
Reach is the estimate of unique people who saw it — always less than or equal to impressions.
Frequency = impressions ÷ reach, the average number of times one person has seen your ad. It's also the earliest fatigue signal — it climbs before cost does.
CPM = (spend ÷ impressions) × 1,000. What you pay to show the ad 1,000 times, regardless of whether anyone clicks.
Why Is "CTR" Always Ambiguous on Meta?
Meta tracks three nested click families, and people quote the wrong one constantly.
- Clicks (all) — every interaction: likes, comments, shares, "see more," profile visits. A vanity metric — ignore it for judging performance.
- Link clicks — clicks toward your actual destination, on or off Meta.
- Outbound clicks — the subset of link clicks that leave Meta entirely. Best for comparing Meta against other channels.
CTR and CPC inherit the same ambiguity. "CTR" alone could mean CTR (all), link CTR, or outbound CTR — three different numbers from one ad. When a media buyer says CTR, they almost always mean link CTR (link click-through rate), and cost per click means CPC (cost per link click).
What's the Difference Between Cost Per Result and CPA?
Cost per result = spend ÷ results, where "result" is whatever the campaign optimizes for — a purchase, a lead, a landing page view or link click for traffic, a ThruPlay for video.
CPA specifically means cost per purchase — spend ÷ purchases.
They're the same number only when your optimized result IS a purchase. Optimize for landing page views, and "cost per result" is cost per LPV, not CPA — report it as CPA and you've just told someone acquisition is cheaper than it is.
How Do Landing Page Views Fit In?
Landing page views (LPV) count only when someone clicked AND the page actually loaded with the pixel firing. Every LPV is a link click, but not every link click becomes an LPV — some people close the tab before the page renders.
A healthy LPV-to-link-click ratio runs roughly 80–90%. Drop below 60% and something specific is broken: a slow-loading page, a misleading ad (the click promised something the page didn't deliver), or targeting pulling in the wrong audience.
| Signal | What it tells you | Where the problem usually lives |
|---|---|---|
| Link CTR low | Ad isn't earning the click | Hook, offer, or creative |
| LPV/link-click ratio < 60% | Click isn't turning into a real visit | Page speed, or ad/page mismatch |
| LPV high, purchases low | Visit isn't converting | Page copy, price, checkout friction |
| Cost per ATC rising, CPA flat | Interest is there, close rate isn't | Product page, shipping cost surprise |
ROAS vs Website Purchase ROAS: What's the Difference?
Purchase ROAS = purchase conversion value ÷ ad spend, counting every conversion in Meta's attribution window — including purchases on Meta itself (Shops) or inside an app.
Website purchase ROAS is narrower: web-only, tracked through the Pixel/Conversions API, excluding on-Meta and in-app purchases.
Sell almost entirely through your site and the two numbers sit close together. Run Shops or an app checkout too, and purchase ROAS reads higher — website purchase ROAS is the one that reconciles better with Shopify or GA4.
Worked example. Say a campaign spent $2,000 and Meta attributes $4,600 in purchase value, of which $3,800 came through your website pixel and $800 came from an on-Meta Shop checkout.
- Purchase ROAS = $4,600 ÷ $2,000 = 2.3x
- Website purchase ROAS = $3,800 ÷ $2,000 = 1.9x
Both numbers are "true" — they're just answering different questions. Report the one that matches how the rest of the business measures revenue.
What Are Hook Rate and Hold Rate, and How Do I Build Them?
Neither is a native Meta column — you build both as custom metrics, and the denominator you pick changes the number.
Hook rate (commonly) = 3-second video plays ÷ impressions × 100. It measures whether the opening of the video earns attention.
Hold rate has two common versions: ThruPlays ÷ 3-second plays × 100 (retention among people who started watching), or ThruPlays ÷ impressions × 100 (retention against everyone reached). Pick one, label it, and stay consistent — mixing the two across reports makes trend lines meaningless.
Benchmark bands vary by source and placement — treat these as tiers, not targets: hook rate under ~20-25% is weak, 25-40% is solid, 45%+ is elite. Hold rate under ~15% is weak; 25-50% is what most sources call good, depending on the denominator used.
How Do I Build a Custom Metric in Ads Manager?
Ads Manager doesn't ship hook rate, hold rate, or cost-per-ATC as ready columns. You build them once, save them as a preset, and reuse them.
- 1Open ColumnsIn Ads Manager, click "Columns" on any table
- 2Customize columnsChoose "Customize columns" at the bottom of that dropdown
- 3Create custom metricBottom-left of the picker, set name + format
- 4Write the formulaUse +, −, ×, ÷ and metric tokens, e.g. spend/purchases
- 5Save as presetTick "Save as preset," name it, apply
One formatting trap catches almost everyone once: if your formula already multiplies by 100 to make a percentage, keep the column Format set to Number, not Percentage. Percentage format multiplies by 100 again — a real 9% turns into a nonsense 900% on screen. For the fuller walkthrough of every preset and pane, see the guide to custom columns in Meta Ads Manager.
What Changed in Meta's Reporting in 2026?
A few dashboard changes are worth knowing so you don't mistake a metric shift for a performance shift.
Opportunity Score is a 0-100 setup score rating how closely your account follows Meta's recommendations. Meta says it doesn't reflect actual or future performance — treat it as a checklist, not a KPI.
View-attribution windows were removed January 12, 2026. The 7-day and 28-day view windows are gone from Ads Manager and the API. A 15-40% overnight drop in reported conversions around that date is a reporting artifact from the window change, not a real performance drop.
Click attribution was narrowed around March 2026 for web and in-store conversion campaigns — only real outbound link clicks count as "click," with other interactions moved into a separate "engage-through" window. The exact default attribution string is reported inconsistently right now; verify it live rather than trusting any article's wording, including this one.
EU/UK Digital Services Tax now shows as a separate invoice line, not folded into Ads Manager spend, so dashboard ROAS and MER understate true cost for EU/UK accounts unless you add that line back in.
How Do These KPIs Roll Up Into Business Metrics?
ROAS in Ads Manager is one input into bigger numbers finance cares about — MER, CAC, contribution margin — each with its own denominator. That fuller calculation is covered in the guide to measuring Facebook ad performance.
Where the Klipio Connection Fits
None of these KPIs exist for a competitor's ads — Meta shows no spend, no CTR, no ROAS for anyone but you. The one public proxy is ad longevity, the "Started running on" date on every Ad Library card. An ad still live after 60-90 days is almost certainly still profitable, because nobody keeps funding a loser that long.
That's the gap our free Meta Ad Library downloader extension and Klipio's paid plans (from $79/mo) close: Klipio watches competitors' live Meta ads, surfaces the longest-running ones, and mines the angles behind them so you're briefing new creative from a validated pattern, not a blank page. It won't hand you their CPA — it shows you which ads Meta's own auction has kept rewarding for months, the closest public proxy available. For more tool options, see the comparison of ad research tools.
FAQ
What are the most important Meta ads KPIs?
For most ecommerce accounts: CPM and frequency for delivery, link CTR and CPC (link click) for engagement, and cost per result plus ROAS for outcome. Which ones matter most depends on your campaign objective and where the funnel is actually breaking.
What's a good CTR for Facebook ads?
It depends which CTR you mean. Median link CTR sits around 1%, while CTR (all) — which includes likes, comments, and shares — runs roughly 1.7-1.8%, close to double. Baseline against your own account's history rather than chasing a fixed number, since it varies heavily by vertical and placement.
Is cost per result the same as CPA?
Only when the campaign's optimized result is a purchase. For a traffic, lead, or engagement campaign, "cost per result" is measuring whatever that campaign optimizes for — check the optimization event before you relabel a number as CPA.
What's the difference between ROAS and website purchase ROAS?
Purchase ROAS counts all attributed conversions, including on-Meta Shop and in-app purchases. Website purchase ROAS counts only conversions tracked through your Pixel/Conversions API on your own site. If you sell only through your website, they should be close; if you also run Shops, purchase ROAS reads higher.
What's a good hook rate on Meta?
Sources disagree, but a rough tiering is: under 20-25% is weak, 25-40% is generally solid, 45%+ is elite. Always check the denominator (impressions vs. reach) before comparing hook rate across sources or accounts.
How often should I check my Meta ads KPIs?
Check delivery metrics like frequency and CPM at least weekly per ad, since fatigue signals appear there before cost per result moves. Review outcome metrics like ROAS and CPA on whatever cadence matches your reporting cycle, but never judge a campaign on engagement metrics alone — a high hook rate with a rising CPA is still a losing ad.
The Klipio extension adds a download button to every ad in the Meta Ad Library: one click per ad, or bulk-save a whole search as a ZIP with a searchable swipe file inside. Free, no sign-up.
Get the free extension
