📋 What you'll need: Your live Meta campaign, access to Meta Ads Manager, and your Creative Tracker (Spreadsheet 3, Results Log tab) open to record what you find
⏱ Time: About 20 minutes a week, starting a few days after your campaign goes live — not before
✅ Output: A clear read on which of your five ads are working, recorded in your Results Log, and a simple keep, cut, or scale decision for each
⏱ Time: About 20 minutes a week, starting a few days after your campaign goes live — not before
✅ Output: A clear read on which of your five ads are working, recorded in your Results Log, and a simple keep, cut, or scale decision for each
Your campaign is live — now the job changes from building to reading. You don't need to watch it every hour, and you shouldn't. Meta needs a few days and some spend before the numbers mean anything. Give it that time, then check in about once a week with one simple question: which ads deserve more money, which deserve none, and which just need a little longer.
This page shows you the handful of numbers worth looking at, what "good" looks like for each, and how to turn them into a decision. Everything you record goes into your Results Log — Spreadsheet 3, Results Log tab.
⚠️ Don't touch it for the first few days.
The most common mistake is judging ads too early — pausing one after a day because it "isn't working." Meta spends the first few days learning who to show each ad to, and results swing around during that window. Give each ad at least 3–4 days and enough spend to gather real data before you decide anything. Reacting to day-one numbers just resets the learning and costs you more.
The Numbers That Matter
Ads Manager shows you dozens of columns. Ignore almost all of them. For a founder testing five ads on a small budget, three numbers tell you what you need — and together they tell a simple story: did the ad stop the scroll, did it earn the click, and did the click turn into a result worth paying for.
Before you start, set up your columns. In Ads Manager, above your results table, click "Columns: Performance" and choose "Customise columns." Add the ones below so everything you need is on one screen. You can save this as your own column preset so it's there every time.
Hook rate — did the ad stop the scroll?
What it is: the share of people who watched the first 3 seconds of your video. It tells you whether your opening stopped the scroll — the most important job any ad has. Video only; static ads don't have a hook rate.
Where to find it: Meta doesn't always show this as a single column, so you may need to calculate it. Add two columns in "Customise columns": "3-second video plays" and "Impressions." Then:
Hook rate = 3-second video plays ÷ impressions × 100
For example, 1,050 three-second plays from 4,200 impressions is a 25% hook rate. Some accounts also have a ready-made "Hook rate" column — search "hook" in the customise columns box and add it if it's there.
What's good: 25%+ is solid, 40%+ is excellent, below ~20% means your opening isn't working.
What to do: a weak hook rate is a hook problem, not a product problem — test a stronger opening using one of the alternative hooks the AI gave you when you built the ad.
Click-through rate (CTR) — did it earn the click?
What it is: the share of people who clicked your ad after seeing it — the bridge between watching and doing.
Where to find it: this one Meta calculates for you. In "Customise columns," add "CTR (link click-through rate)." Use the link click version, not "CTR (all)," which counts likes and shares you don't care about here. If you'd rather check it yourself:
CTR = link clicks ÷ impressions × 100
What's good: 1%+ is a reasonable starting point, well above 1% is strong, well below 1% means a mismatch somewhere.
What to do: read it against the hook rate. A strong hook rate but weak CTR means people watch but aren't compelled to click — usually a message or offer problem. For a skincare brand, that might mean the opening grabs attention but the offer isn't strong enough — worth testing a first-order discount.
Cost per result — what you pay for a lead or sale
What it is: what you pay, on average, for one real outcome — one lead (service) or one sale (ecommerce). This is the number that decides everything.
Where to find it: Meta shows this natively. In "Customise columns," make sure "Results" and "Cost per result" are both added — they'll be tied to the conversion event you chose when you built the campaign (a lead, a purchase). If cost per result looks blank or wrong, it usually means the conversion event isn't set up correctly — check back in your ad set, or revisit SOP 1.
What's good: there's no universal number — it depends entirely on what a lead or sale is worth to you, which is why you worked out your numbers on the Know Your Numbers page. A financial advisor who knows a new client is worth €2,000 can afford a lot per lead; a candle maker selling a €25 candle needs their cost per sale much lower. Compare what you're paying against what you can afford: paying €11 for a lead you value at €20 is working; paying €40 for it is losing money.
What to do: this is the number your keep, cut, or scale decisions are built on. Hook rate and CTR explain why an ad is or isn't working; cost per result tells you whether it's working at all.
💡 Read the three as one story.
- Hook rate tells you if the ad stopped the scroll
- CTR tells you if it earned the click
- Cost per result tells you if the click paid off.
When an ad underperforms, the chain shows you where to look: weak hook rate → fix the opening; good hook rate but weak CTR → fix the message or offer; strong on both but high cost per result → look at your landing page or how well the audience is matching. You're following a trail to what needs fixing.
Keep, Cut, or Scale — Your Weekly Decision
Once a week, open your Results Log (Spreadsheet 3, Results Log tab), fill in the numbers for each of your five ads, and make one decision per ad. There are only three: keep it running, cut it, or put more budget behind it. That's the whole job — you're not rewriting ads every day, you're reading the numbers once a week and moving budget toward what works.
Keep — give it more time
For ads that are close but not yet clear: a decent hook rate and CTR, a cost per result near what you can afford, but not enough spend behind them yet to be sure. Leave these running as they are and check again next week. Not every ad declares itself quickly — some just need a little more data before the picture is clear. When in doubt and the ad isn't losing money, keep it.
Cut — stop the losers
For ads that have had a fair run — at least 3–4 days and enough spend to judge — and are clearly underperforming: a low hook rate, a weak CTR, and a cost per result well above what you can afford. Turn these off. This is your first and most important move each week, and it does two jobs at once.
It stops the waste — every euro a losing ad spends is one that could go to a winner. And because your budget sits at the campaign level, cutting a loser doesn't just save that money, it redirects it: Meta pushes the freed-up spend straight to the ads that are working. So turning off your losers is the simplest, fastest way to put more behind your winners, without touching anything else.
Cutting isn't giving up — it's the whole point of testing five ads. You expected some to lose. A candle maker who finds two of their five ads costing €30 per sale on a €25 candle isn't failing — they're doing exactly what testing is for: finding the losers fast and cutting them so the winners get more room.
Scale — grow what's working
Once you've cut the losers, Meta is already concentrating your budget on the winners. The next move — and only once the campaign is profitable overall, with your cost per result comfortably below what you can afford — is to grow the whole campaign budget.
Raise it by no more than about 20% every few days. Because Meta is already favouring your best ads, that extra budget flows mostly to them. Raise it too fast, though, and you reset Meta's learning, which can make a winning campaign wobble — slow and steady keeps it stable.
Keep challengers coming: when you have a clear winner, produce a fresh wave from your leftover concepts and test them against it, so something new is always trying to beat your best. A personal trainer whose "still sore three days later?" ad is bringing in leads at €9 should be building the next three to try and beat it, not sitting still.
Give a proven winner its own campaign: Once an ad has performed consistently for a couple of weeks, take it out of your test campaign and give it its own campaign with its own budget. This does two things: it lets you put real budget behind your winner without it competing against your other ads for spend, and it protects it from being disturbed every time you add new tests. Your original campaign stays your testing ground; your winners graduate into their own campaigns where they can grow steadily. This is how a single winning ad becomes the engine of your account.
Watch for Creative Fatigue
Here's something that happens to every winning ad eventually: it slowly stops working. Not because the ad got worse, but because the same people have seen it too many times. The scroll-stopping hook they saw once is now the ad they've scrolled past five times. This is creative fatigue, and spotting it early is what separates a campaign that keeps performing from one that quietly dies.
Frequency — how to spot fatigue coming
What it is: frequency is the average number of times each person has seen your ad. A frequency of 1 means the average person has seen it once; 3 means three times. It's the clearest early signal that fatigue is setting in.
Where to find it: in Ads Manager, add the "Frequency" column through "Customise columns" (search "frequency"). Check it when you do your weekly review.
What's good, what's a warning: on a small, cold audience there's no hard rule, but as frequency creeps toward 2.5–3 and keeps climbing, watch your other numbers closely. Fatigue usually shows up as frequency rising while your CTR slowly falls and your cost per result slowly climbs — the same ad, the same audience, just worn out from overexposure. On a small budget this happens faster than you'd think, because you cycle through the same limited pool of people quickly.
What to do: don't wait for a winning ad to collapse. When you see frequency climbing and performance slipping, it's time to refresh — put new creative in front of that audience before the old ad burns out completely.
💡 This is exactly why you kept the rest.
Remember the concepts you didn't pick for your first five? This is when they earn their place. The habit that keeps a campaign alive is refreshing your creative every few weeks — producing a fresh wave from your leftover concepts and adding it in before fatigue sets in. A skincare brand whose best ad has been running six weeks shouldn't wait for it to die — they should already have the next batch ready to swap in. You're never relying on one ad forever; you're always cycling in something new.
When Something Isn't Working
When an ad underperforms, don't guess at the fix. The three numbers you already track point you straight to the problem — read them in order and each one tells you where to look next. Most of the time the fix is one specific thing, not a rebuild.
Weak hook rate → fix the opening
If the hook rate is low (below about 20%), people are scrolling past before the ad ever gets going. The problem is your first few seconds — nothing else, because most people never see the rest.
The fix: keep the concept, change the opening. Test one of the alternative hooks the AI gave you when you built the ad, or write a sharper one. A therapist whose ad opens with "At my practice, we believe in a holistic approach" is losing people instantly — swapping to "Lying awake at 3am with your mind racing?" gives the right person a reason to stop. Change only the first few seconds and run it again.
Good hook rate, weak CTR → fix the message or offer
If people are watching (hook rate is fine) but not clicking (CTR below about 1%), the opening works but what follows isn't compelling enough to make them act. The problem is in the body, the offer, or the match between the ad and who it's reaching.
The fix: look at what you're actually asking people to do and whether it's worth it to them. Is the offer clear and worth clicking for? Does the middle of the ad build on the hook or wander off? A candle maker whose ad grabs attention with a beautiful pour shot but asks people to "explore our collection" might do far better offering "15% off your first order" — a concrete reason to click now, not later.
Good hook and CTR, high cost per result → look past the ad
If people are watching and clicking, but your cost per result is still too high, the ad is doing its job — the problem is after the click. The two usual suspects are your landing page and the match between your ad and your audience.
The fix: check that the page people land on delivers what the ad promised. If your ad sells "15% off your first order" but the landing page opens with your brand story and buries the offer, people leave — and you've paid for the click for nothing. Make sure the page picks up exactly where the ad left off: same promise, same offer, immediately visible. If the page is solid and the cost is still high, the ad may simply be reaching the wrong people — which is your cue to test a genuinely different angle that pulls a different audience.
💡 One change at a time.
When you fix something, change only that one thing and run it again. If you rewrite the hook, swap the offer, and change the landing page all at once and the ad improves, you'll never know which change did it — so you can't repeat it. Slow, single changes are how you learn what actually works for your business.
You've Done Something Most Founders Never Do
Step back and see what you've built. You researched your buyer, made five genuinely different ads, launched a real campaign, and you can now read what it's doing and improve it week by week. That's a complete paid traffic system — the exact thing most small founders pay agencies thousands to run, and it's yours, running in your business, understood by you.
From here, the work is rhythm, not reinvention. Once a week: read your numbers, cut the losers, feed the winners, watch for fatigue, and cycle in fresh creative from the concepts you've already got waiting. That loop is how a small budget compounds into steady, predictable results over time.
💡 Where to next?
If you want a second channel, Module 4 shows you how to reach buyers at the exact moment they search for what you offer with Google Ads — a natural complement to the demand you're now creating on Meta. If Meta is enough for now, carry on to Module 5. Either way, keep this campaign running and keep working the weekly loop — momentum is everything from here.
Made with Bullet