The short answer: create one custom column set with six metrics and use it for every review: cost per thousand impressions (CPM), outbound click-through rate, cost per outbound click, landing page conversion rate, cost per result and frequency. Read them as a chain from left to right, because the first metric that looks wrong points to where the problem is. Hide everything else until you have a specific reason to look at it.
Meta’s Ads Manager is generous with columns. Reach, impressions, post reactions, post saves, video plays, page engagement and dozens more are one click away. Most of them are context, not decisions. When we audit a new client account, the most common finding is not a technical error but a dashboard so crowded that nobody could see what was wrong.
Why does the default view mislead?
The default columns emphasise things Meta can easily count: reach, impressions, results and amount spent. Those are fine, but they leave out the metrics that explain why cost per result moved. A rising cost per lead can come from four different causes: more expensive impressions, a weaker ad, a weaker landing page or a tired audience. Each demands a different fix, and the default view does not separate them.
A custom column set, saved once and reused, solves that.
The six metrics, read as a chain
1. CPM: what does reaching 1,000 people cost?
CPM is the price of attention in the auction. It reflects your audience, the season, competition and how Meta rates the quality of your ad. A sudden jump is often seasonal: we cover Q4 patterns in our piece on how much CPMs actually rise. A high CPM on its own is not bad if the people reached convert well; it is a cost input, not a verdict.
2. Outbound click-through rate: does the ad make people leave Meta?
Use the outbound (link) click-through rate, not the all-clicks version. “Clicks (all)” includes likes, comments, profile taps and expanding the text, so it flatters weak ads. Outbound CTR counts only people who clicked through to your destination. If CPM is steady but outbound CTR falls, the creative or the offer is the likely cause.
3. Cost per outbound click: what does a visit cost?
This combines the first two: a cheap CPM with a low CTR can still produce an expensive click. Track it as a sanity check, but do not optimise it in isolation. A cheap click from the wrong audience is still waste.
4. Landing page conversion rate: what happens after the click?
Ads Manager does not always show this neatly. Calculate it as results divided by outbound clicks (or landing page views), or take it from your analytics. If clicks are cheap and plentiful but few turn into results, the problem is probably after the ad: slow page, mismatch between ad and page, unclear offer or a form that asks too much. Our article on lead form ads versus landing page ads explains one structural fix.
Note that landing page views is a better denominator than clicks, because some clicks never load the page. If the gap between the two is large, check page speed and the tracking pixel.
5. Cost per result: what does an outcome cost?
This is the number you ultimately care about, but only if the “result” is the right one. Confirm which event Ads Manager counts as a result for each campaign (lead, purchase, add to cart, landing page view). A campaign optimised for landing page views will show beautifully low costs per result and bring no customers. Compare cost per result with your break-even point; our guide to break-even ROAS shows how to calculate it.
6. Frequency: how often has the same person seen the ad?
Frequency is average impressions per person. As it climbs, performance on the same audience usually deteriorates because people ignore or dislike a repeated ad. There is no universal limit. For cold audiences, rising frequency combined with falling CTR is the warning sign; for small retargeting audiences, higher frequency is normal and can be fine for a few days. Our guide to the ad fatigue playbook covers what to do when you see it.
How do you use the chain to diagnose a problem?
When cost per result worsens, walk the chain from left to right and stop at the first metric that changed:
- CPM up, the rest stable: the auction is more expensive. Check season, audience size and overlapping campaigns. Consider waiting or broadening the audience.
- CTR down, CPM stable: the creative is weaker or the audience is tired. Test new creative.
- CTR fine, conversion rate down: the landing page, offer or tracking changed. Look at the page.
- Frequency high and CTR falling: saturation. Refresh creative or widen the audience.
This takes five minutes and prevents the most expensive mistake, which is changing everything at once so that you never learn which change mattered.
What about video metrics?
For video ads, add two more columns temporarily: 3-second video plays and ThruPlays (the plays that reached 15 seconds or the end, depending on length). Divide 3-second plays by impressions to get what many advertisers call the hook rate: the share of people who stopped scrolling long enough to watch. Compare ThruPlays with 3-second plays to see how many stayed. A low hook rate means the opening needs work; a good hook rate with weak holding means the middle drags. These are our working conventions, since Meta does not label “hook rate” as a standard column. See our guide to briefing a video editor for paid social for how to fix openings.
Which columns can you safely hide?
- Likes, reactions, comments, shares and saves: useful for social proof, not for lead or sales performance.
- Reach and impressions in isolation: context for CPM and frequency, not results.
- Clicks (all): replace with outbound clicks.
- Page likes and follows: unless that is your actual goal.
- Quality, engagement and conversion ranking: relative diagnostics against competing ads, which can help occasionally but should not drive decisions.
You can keep them available in a separate “extended” column set for deep dives.
How do you set this up?
- In Ads Manager, open the Columns dropdown and choose Customize Columns.
- Select amount spent, CPM, outbound CTR, cost per outbound click, results, cost per result and frequency. Add landing page views if available.
- Save it as a preset and set it as default for your account.
- Set breakdowns to Day or Week for trend, and to Placement or Age occasionally for diagnosis.
- Choose attribution settings and keep them constant when comparing periods, since changing them changes the numbers.
How often should you look?
For most small accounts, a short review twice a week and a longer one weekly is enough. Daily checking mostly produces unnecessary edits, which can restart the learning phase. Our post on when to pause a Meta campaign and when to let it run offers a decision rule for the most common reaction to a bad day.
A simple dashboard of six numbers will not make an average ad good. What it will do is tell you, in a few minutes, which part of the system to fix, and that is most of what useful reporting does.
A weekly five-minute routine
Once the column set is saved, a short routine keeps you honest. Set the date range to the last seven days and compare it with the previous seven. Sort campaigns by amount spent so that the biggest budget gets the first look. For each of the top three, walk the chain: CPM, outbound CTR, conversion rate, cost per result, frequency. Write one line in a notes document: what moved, what you think caused it and what you will do, including “nothing” where that is the right call. After a month those notes form a record of cause and effect that is more valuable than any single dashboard. They also make it easy to hand the account to someone else, or to explain to an agency or a colleague why a decision was made.
Avoid judging ad sets or ads that have had very little spend. As a rough guide, wait until an ad has produced several results, or has spent a few times your target cost per result, before drawing conclusions. Small samples produce dramatic percentages that vanish a day later.