You will judge real ads by what the auction is likely to reward.
You have probably clicked an ad that promised one thing and landed on a page about something else. A sale banner that led to full-price items. A free trial that turned out to need a credit card and a phone call. You left, and the advertiser paid for your click anyway. Platforms see that pattern across millions of visits, and it feeds the quality signals you met in lesson 1.1, The highest bid does not always win.
This exercise trains your eye before you spend anything. You will look at real ads that are running right now, judge them the way an auction is likely to, and decide what you would fix. It is easier to spot problems in someone else's ad than in your own, and the habit carries over.
Three public tools let anyone look at ads without logging in to an ad account.
The Meta Ad Library lets you search by keyword or by the name of a Facebook Page, filter by country, and see the active ads that page is running on Facebook and Instagram. The Google Ads Transparency Center lets you search for an advertiser by name or website and filter by region, showing search, display and video ads they have run. The TikTok Creative Center has a section on top ads that you can filter by region and industry; it shows a selection of ads that performed well rather than every ad running, which still makes it useful for seeing what holds attention.
Set the country or region to Singapore where you can, and search for businesses in your own category. Pick three ads that a buyer like yours would plausibly see.
For each ad, write one sentence about who is likely to see it and what they want at that moment. For a search ad, start from the search it probably appeared on. For a feed or video ad, start from the kind of person the advertiser seems to be targeting.
This matters because relevance is always relative to a person. An ad can be clear and well made and still be irrelevant to the person in front of it.
Give each ad two scores from 1 to 5. The scale is just a working tool for this exercise, not anything the platforms use.
Relevance asks whether the ad speaks to what that likely viewer wants. Does it use their words, name their situation and make an offer that fits?
Landing page match asks whether the page delivers what the ad promised. Click through. Does the headline repeat the promise? Is the price, trial or product the ad mentioned easy to find on the first screen, on a phone?
As you go, note signs of low quality that platforms say they look for: clickbait headlines that hold back the point, prices or key terms withheld until later, exaggerated claims, and a gap between ad and page.
Last, for each ad, write the one change you would make first if it were yours. One change, not five. Choose the one most likely to improve how both the viewer and the auction treat the ad.
Priya runs a maths centre in Tampines. She searched the Google Ads Transparency Center and the Meta Ad Library for tuition centres in her area and picked three ads. The scores below are her own judgements.
The first was a Google search ad from a large chain, probably showing on searches like maths tuition Tampines. The headline mentioned maths tuition and several locations, and the description listed every subject from English to science. The link went to the homepage, which had a carousel of photos and a menu of eight programmes. Priya gave it 3 for relevance and 2 for landing page match. A parent looking for P5 maths would have to dig. Her first fix: send the click to the chain's maths page for the Tampines branch.
The second was a Meta ad from a small centre in Pasir Ris. It showed a short clip of a tutor explaining a model drawing question, with the words free trial class for P5 and P6 maths. The link went to a page with the same headline, a timetable and a short form. She gave it 5 for relevance and 4 for match, taking one point off because the form asked for the child's school and both parents' details, which she thought was too much at that stage. Her first fix: cut the form down to name, phone and level.
The third was another Meta ad, with the headline you won't believe what this student scored over a photo of a smiling child holding a certificate. The page talked about the centre's philosophy and listed no prices or class times. She gave it 2 for relevance and 1 for match, and noted it as clickbait with withheld information, both things Meta says its ad quality systems consider. Her first fix: replace the headline with a plain offer and put fees and timetables on the page.
Looking at the three together, Priya noticed something uncomfortable. Her own ads looked more like the first one than the second. That was the point of the exercise.
When you are done, you should have three short entries. Each one names the ad, the likely viewer, the two scores, any quality warning signs, and one fix. Keep it to a page. You will use the same eye later when you review your own drafts in module five, so keep the notes somewhere you can find them.
Open the three libraries now and start with your own category. In the activity below you will record your three ads, their scores and the first fix for each.
Pick three live ads in your category from the public ad libraries, score each on relevance and landing page match, and write one fix for each.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).