Display Advertising: When It Works, When It Does Not
The cheapest impressions in digital, and the easiest budget to waste. Both facts are true, and the difference between them is about six settings.
Display has a reputation problem, and it is half deserved.
Run cold, broadly targeted, with default settings, it will spend a budget and return very little. Plenty of people have tried exactly that and concluded the channel is dead.
Run as remarketing, or tightly targeted with the obvious waste excluded, it does a job nothing else does at the same price.
The difference is not the channel. It is about six settings and one measurement decision.
What display is for
Search captures demand that already exists. Someone types what they want and you appear.
Display interrupts. Nobody browsing a news site was looking for you. That makes it poor at converting cold traffic and useful for two other jobs.
| Job | Does display do it well? | Why |
|---|---|---|
| Remarketing | Yes — the strongest use | The audience already knows you. You are reminding, not introducing. |
| Awareness in a defined audience | Yes, if measured properly | Cheap reach at scale, when the targeting is genuinely narrow |
| Cold direct response | Rarely | You are interrupting someone mid-task with an unfamiliar brand |
| Launching a new category | Sometimes | Where nobody is searching yet, there is no query to buy |
| Competing on high-intent commercial terms | No | That demand belongs to search |
Where the money actually goes
Almost every underperforming display campaign has the same problem, and it is not the creative.
Mobile app placements. Your ad appears inside a game or a utility app. Someone tries to close it, misses, and taps the ad. That click costs you real money and will never convert. At scale it can consume a large share of a display budget while producing a click-through rate that looks encouraging.
Exclude mobile app categories before you launch. This one change fixes more display campaigns than everything else combined.
After that, in order:
- Read the placement report weekly. Sort by cost. Exclude anything obviously irrelevant — you will find sites you would never have chosen.
- Cap frequency. Without a cap, a small audience sees your ad many times a day. That is not persuasion, it is attrition.
- Narrow the audience. Broad display targeting is the default and the weakest option. Remarketing lists, customer lists and tightly defined in-market segments all beat it.
- Exclude converters. Same discipline as remarketing. The ad following someone around after they bought is the reason people dislike this channel.
- Check content exclusions. Decide what your brand should not appear next to, and enforce it.
- Build a shared exclusion list. Apply it account-wide so every new campaign inherits what you have learned.
Google Display Ads specifically
Google's display inventory reaches an enormous share of the web, which is both the appeal and the risk. Scale without exclusions is just a faster way to spend.
Targeting options worth using, roughly in order of reliability: your own remarketing and customer lists first, then in-market segments where the definition genuinely matches your buyer, then custom segments built from competitor sites or relevant search behaviour. Broad demographic targeting last, if at all.
Note that Performance Max also buys display inventory, automatically and without placement-level control. If you are running both, decide deliberately which owns the display budget rather than letting them compete inside your own account.
Creative that survives being ignored
Display creative is seen peripherally, at small sizes, by someone doing something else. That constrains it more than most formats.
- One message per ad. There is no room for two, and no attention for two.
- Legible at the smallest size you run. If the headline is unreadable on a mobile banner, it does not exist.
- Brand visible immediately. Unlike search, nobody asked who you are, so an unbranded ad wastes the impression entirely.
- Contrast over subtlety. This is peripheral vision territory.
- Build all the sizes. Missing sizes silently limit where you can serve.
- Refresh before fatigue. Display fatigues fast because frequency accumulates quickly.
How to measure it without fooling yourself
Judging display on last-click conversions is how it gets cancelled unfairly.
Someone sees a display ad, does nothing, and searches your brand three days later. Attributed to last click, that conversion belongs to search. Display looks like it did nothing.
Track instead:
- Branded search volume before, during and after the flight. Free to check in Search Console and the cleanest available signal.
- Assisted conversions rather than last-click revenue.
- Whether remarketing lists built from display viewers convert better than cold ones.
- Blended cost per acquisition across the whole account over a period long enough to include the lag.
The same attribution problem applies to YouTube, for the same reason. Both create demand that search then captures.
Frequently asked questions
Does display advertising still work?
For remarketing and for building awareness in a defined audience, yes. As a way to generate cold direct-response conversions, rarely — you are interrupting someone who was reading something else. Judge display on assisted conversions and branded search lift rather than last-click revenue, or it will always look like a failure.
Why is my display campaign getting clicks but no conversions?
Usually mobile app placements. Ads inside games and utility apps collect accidental taps from people who were trying to close them. Those clicks cost real money and never convert. Excluding mobile app categories is the first fix in almost every underperforming display campaign.
What is the difference between display advertising and Google Display Ads?
Display advertising is the category — visual ads shown on websites and apps. Google Display Ads is Google's version of it, running on the Google Display Network. Meta's feed placements are also display in the broad sense, but the industry usually means the open web when it says display.
How do you stop wasting money on display?
Exclude mobile app placements, review the placement report weekly and exclude anything irrelevant, cap frequency, and set a tight audience rather than relying on broad targeting. Most display waste is a handful of placements consuming a large share of budget, which is why the placement report matters more than the bid.
Should display be measured on last-click conversions?
No, and doing so is why display gets cancelled unfairly. Someone sees a display ad, does nothing, and searches your brand three days later. Last click gives that conversion to search. Track assisted conversions and branded search volume across the flight instead.
We cut the waste before we raise the budget
Placement audits, app exclusions, frequency discipline, and measurement that credits display for what it actually caused.