Quick answer: Read Facebook ads results in order: real leads and cost per lead first, then reach, CPM, link CTR, landing page views and page conversion rate.
Ads Manager can show you more than 300 metrics. For a service business, about eight of them matter, and they matter in a specific order. Reading them in the wrong order is how people end up “fixing” the wrong thing, like changing the audience when the real problem is the landing page.
Start with the result that pays the bills
Before looking at anything else, answer one question: how many real enquiries did we get, and what did each one cost?
In Ads Manager that means two columns:
- Results: the number of times your chosen optimisation event happened (lead, form submission, purchase)
- Cost per result: spend divided by results
Then check those numbers against reality. Count the enquiries in your inbox or CRM for the same period. If Meta says 40 leads and you received 22, your tracking is over-counting (or counting something other than real enquiries). If Meta says 10 and you received 30, it’s under-counting. Fix that first. Nothing else is reliable until it matches. See why your ad numbers don’t match.
Then work down the funnel
If results are poor, find where people drop off. Read these in order:
1. Reach and frequency
- Reach: how many different people saw the ad
- Frequency: average times each person saw it
A frequency above about 3 to 5 in a week on a cold audience usually means the audience is too small or the ad has worn out. See Facebook ad fatigue.
2. CPM (cost per 1,000 impressions)
What you pay for attention. CPMs vary by season, audience and competition. Australian CPMs commonly sit somewhere in the $10 to $30 range, rising sharply in November and December. A very high CPM can point to a tiny audience, a special ad category, or poor engagement with the ad itself.
3. Link click-through rate (CTR, link)
Use CTR (link click-through rate), not plain “CTR (all)”, which counts likes, profile clicks and “see more” taps. For service ads, a link CTR around 1% or more is healthy. Well below that and the creative or offer isn’t stopping people.
4. Cost per link click and landing page views
If clicks are cheap but landing page views are much lower than link clicks, people are bouncing before your page loads. That usually means a slow page or accidental clicks from certain placements.
5. Conversion rate on the page
Results divided by landing page views. If people arrive but don’t enquire, the problem is the page or the offer, not the ad. A reasonable service landing page converts somewhere between 3% and 10% of visitors. See why Facebook ads get clicks but no enquiries.
Understand attribution before you trust “Results”
Meta counts a result when it happens within a set window after someone interacts with your ad. The default for website conversions in 2026 is:
- 7-day click: they clicked your ad and converted within 7 days
- 1-day engage-through: they liked, commented, shared or watched (rather than clicked) and converted within a day
- 1-day view: they saw the ad, didn’t interact, and converted within a day
View-through and engage-through results are real signals, but they’re also where over-crediting happens. Someone who saw your ad on Instagram and then Googled you and enquired may be counted by both Meta and Google. Use the Compare attribution settings option in the columns menu to see how many of your results come from clicks alone.
For lead generation, the 7-day click number is usually the most honest figure to compare against your CRM.
Metrics you can mostly ignore
- Post engagement and reactions. Nice, but not a business result.
- CTR (all). Inflated by non-link clicks.
- Relevance or quality rankings in isolation. Useful hints, not goals.
- “Estimated” columns such as estimated ad recall lift, unless you’re running awareness campaigns.
How often to check
Daily checks lead to daily changes, and daily changes keep campaigns stuck in the learning phase. A sensible rhythm:
- Daily: glance for anything broken (rejected ads, spend stopped, a sudden spike)
- Weekly: results, cost per result, frequency, CTR, landing page conversion rate
- Monthly: cost per lead against cost per sale, and whether the leads became paying customers
That last step matters most. A $20 lead that never answers the phone is more expensive than a $60 lead that books. See cost per lead explained.
Sources and further reading
- Jon Loomer: How Meta ads attribution works in 2026
- Meta Business Help Centre – search “about attribution settings” and “about metrics”
- Meta: About the learning phase
Written by Gregory
20 years building websites and running lead generation, mostly with my own money on the line. Based in Sydney, working with clients worldwide. How I work →