You will be able to read a campaign report and make one clear decision from it.
Open any ad report and you get a wall of columns: impressions, reach, frequency, clicks, click rate, cost per click, results, cost per result, and more if you go looking. Faced with all of it, most people either stare at the one number that looks worst or give up and go by feel. Neither leads to a good decision.
You only need to read a report in a set order, from the outcome back to its causes. Done that way, a report usually points to one clear change.
The first number to look at is cost per result: what you paid for each conversion the campaign was set up to get. Put it next to your target from lesson 8.1, Start from what a customer is worth. Is it under, near or over?
That single comparison tells you whether there is a problem at all. If cost per result is comfortably under target, the campaign is working, and the question is whether to scale it, using lesson 9.2, Scale what works without breaking it. If it is over, the rest of the report helps you find out why.
Cost per result breaks into two parts. Cost per result equals cost per click divided by the conversion rate, the share of clicks that became results. A campaign can be expensive because clicks cost too much, because too few clicks convert, or both. The next two numbers tell you which.
Click rate is the share of people who saw the ad and clicked. A low click rate usually means the ad is not stopping the right people. The creative is weak, the hook does not land, or the ad is reaching people it does not speak to. That points back to module five, on creative, and module four, on audiences. A low click rate also tends to raise cost per click, because the platforms rate ads people ignore less favourably.
Conversion rate is the share of clicks that became results. A low conversion rate usually means something goes wrong after the click. The landing page does not keep the ad's promise, is slow or confusing on a phone, or the offer itself is not attractive enough once people see the details. That points to lesson 5.4, The landing page finishes the ad's job, and lesson 5.3, One offer, one proof, one action.
Here is a sample report from Farah's shop, with two ad sets that each spent S$420. These are example figures.
Ad set A showed the ad 30,000 times and got 450 clicks, a click rate of 1.5 percent, at about S$0.93 a click. Nine of those clicks became sales, a conversion rate of 2 percent. Cost per sale: about S$46.67.
Ad set B showed the ad 32,000 times and got 960 clicks, a click rate of 3 percent, at about S$0.44 a click. Eight became sales, a conversion rate of about 0.8 percent. Cost per sale: S$52.50.
At first glance, B looks better. Twice the click rate, half the cost per click. But it sold less, at a higher cost per sale. Its ad is good at getting clicks, and the clicks are not turning into orders. When Farah opened both ads, she saw why. Ad B promised a weekend sale, and the page it led to showed full prices with the discount only applied at checkout. Her one change: fix ad B's landing page to show the sale prices at the top, and leave everything else alone.
The platform's count of results and your own sales records will rarely agree exactly. The platform may miss conversions it cannot see, as lesson 3.2, Server-side events and first-party matching, explained. It may also count conversions from people who saw an ad and bought days later, depending on its attribution settings, even if they also found you some other way. Your own records miss which ad, if any, a customer saw.
Check both. Compare the platform's results with orders, bookings or enquiries in your own system for the same period. A gap is normal. A gap that suddenly widens, or results reported with no matching orders at all, usually means tracking has broken or is double-counting, and that needs fixing before you trust anything else in the report.
How to share credit for a sale between ads, search, email and word of mouth is called attribution, and it is a subject of its own. It is taught in Marketing analytics: GA4, attribution and testing. For this course, the practical rule is to judge each campaign on its own reported cost per result, sanity-checked against your real sales.
End every review with one decision and the number that justifies it. Not a list of ten tweaks, which would reset learning and make the next report unreadable. One change: pause an ad set, fix a page, replace a creative, raise a budget a step. Write the number next to it, such as conversion rate of 0.8 percent against 2 percent on the matching ad set.
In the activity below you will take a report from a past or sample campaign, work through it in this order and write the one change you would make, with the number behind it.
Take a report from a past or sample campaign and write the one change you would make, with the number that justifies it.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).