Google Display Ads Managed With Placement Discipline

Why do Google Display Ads get such a bad reputation?

Because a default display campaign will spend a meaningful share of its budget on mobile app inventory, where most clicks are accidental. Nothing about that is hidden — it just requires knowing to exclude it before launch rather than after the first invoice.

The second cause is measurement. Display generates impressions cheaply, so cost-per-click and cost-per-impression look excellent next to search. If those are the headline metrics, a campaign delivering nothing can look like the best performer in the account.

Run with exclusions, frequency caps and honest attribution, display does a real job. Run on defaults, it is the fastest way to spend a budget with nothing to show for it.

Should display be judged on direct conversions?

Usually not, and pretending otherwise leads to bad decisions in both directions. Display interrupts people who were not looking for you; expecting the same conversion behavior as a search click misreads the channel.

Where display genuinely earns its place is retargeting people who already visited, supporting a launch or campaign with reach, and building familiarity in a category where the purchase takes weeks of consideration.

That means view-through conversions need care. They count someone who saw an ad and later converted without clicking. Some of that influence is real. Presenting all of it as attributable return is how display campaigns get credited for sales that search earned.

What is the difference between display and paid social?

Both interrupt rather than intercept, but the targeting and the creative expectations differ substantially.

Display places ads around content people are reading, so the ad is peripheral and the creative has to work at a glance in a small slot. Paid social places ads inside a feed people are actively scrolling, where the creative competes with the content itself and native, well-made assets outperform anything that looks like an ad.

Targeting differs too. Social platforms hold richer first-party behavioral data; display relies more on browsing context, in-market signals and your own remarketing lists. Neither replaces the other, and businesses that need reach usually benefit from both weighted differently.

How many creative sizes do we actually need?

Fewer than the full list, more than one. Responsive display ads cover most inventory automatically, but a handful of fixed sizes still carry a disproportionate share of quality placements.

The practical approach is responsive assets as the base, plus dedicated versions of the sizes your placement report shows are actually serving. Building every size in the spec before you know where the ads run is work done in the wrong order.

Creative fatigue is the other half. The same asset shown repeatedly stops working, and on display that happens faster than most teams expect. Planning for rotation from the start is cheaper than diagnosing a performance drop three months in.

How do you stop display wasting money?

By treating placement management as ongoing work rather than a setup task. Left alone, campaigns drift toward the cheapest inventory available, and the cheapest inventory is cheap for reasons.

The first defense is exclusions applied before launch — mobile app inventory unless you specifically want it, obviously irrelevant categories, and the content types your brand should not appear beside. This takes minutes and prevents a substantial share of typical waste.

The second is reading the placement report on a schedule and excluding what does not belong. This is unglamorous and it is where most of the value in display management sits.

The third is frequency capping. Without it a small audience can see the same creative many times a day, which converts nobody and actively damages how people feel about the brand.

How should display be judged?

Not on last-click conversions alone, and not on view-through conversions taken at face value. Both are misleading in opposite directions.

Last-click understates display badly, because display rarely closes a sale. Someone who saw an ad, searched a week later and converted is recorded entirely to search, and a purely last-click view will conclude display did nothing.

View-through overstates it just as badly. Counting a conversion because an ad was technically on a page someone loaded, without a click, attributes a great deal of activity that would have happened anyway.

The honest approach is to look at both, treat neither as the answer, and where budget allows run a holdout — a comparable audience not shown the ads — which is the only method that answers the question directly rather than by inference.

How many creative sizes do you actually need?

Enough to cover the inventory that matters, which is a smaller set than the full list of available formats.

A handful of common sizes account for most available placements. Producing every possible format is a real cost and buys access to a shrinking amount of additional inventory, most of it low value.

Responsive display ads change the calculation. Supplying assets and letting the system assemble combinations covers far more placements for far less production work, at the cost of precise control over how any individual ad looks.

What matters more than count is whether the creative works at the size it will be seen. A carefully composed design becomes an unreadable smear at small sizes, and testing each asset at its actual dimensions is a check that gets skipped surprisingly often.

What is the difference between display and paid social?

Where the person is and what they are doing. Display reaches people reading something else across a network of sites. Paid social reaches people inside a feed they chose to open.

That changes the creative entirely. Display competes with content for attention and has a fraction of a second to register, which favors simplicity and brand recognition. Social sits inside content and can hold attention longer, which favors something worth watching.

Targeting differs too. Social platforms know a great deal about their users from their own behavior; display targeting is assembled from browsing signals and is generally less precise, particularly as third-party cookie support continues to narrow.

Neither is better in general. Display tends to be cheaper for reach; social tends to be stronger for engagement and for creative that needs more than a moment. Most businesses running both should be judging them on different objectives — and social advertising covers that side directly.

How should audiences be built for display?

From your own data first, and from platform-inferred interests only where your own data cannot reach far enough.

Remarketing audiences built from site behavior are the strongest, because the signal is something the person actually did rather than something inferred about them. Segmenting those by what they looked at, and how recently, is where most of the available precision is.

Customer lists are the second. Uploaded with proper consent, they support both direct targeting and lookalike expansion, and they are less affected by browser-level tracking restrictions than pixel-based audiences.

Platform interest and in-market audiences are broader and less certain. They are useful for reaching people who have never visited, and they should be judged against a lower expectation than remarketing rather than compared directly with it.

What is worth avoiding is stacking many narrow audience criteria together. Each addition compounds the inference error, and the resulting audience is frequently both small and less accurate than a broader one would have been.

How should creative be built for display?

For a fraction of a second of attention, in a space competing with the content someone actually came for.

That constraint rules out most of what looks good in a design review. Long copy, subtle typography and detail that rewards study are wasted, because nobody studies a display ad. What registers is one clear message and a recognizable identity.

The brand needs to be visible even when nothing else is absorbed. Someone who does not click should still have registered who it was, because that is a large share of what display is actually delivering.

Legibility at the smallest size is the practical test. A design composed at a large size and scaled down becomes an unreadable smear at the dimensions most placements use, and checking each asset at its actual size catches this before spend does.

And there needs to be more than one. Creative fatigue is a real and measurable decline, and an account with a single asset has no answer to it beyond reducing spend.