A standalone Display campaign
You choose the audiences, hold the exclusion lists, set frequency, and read placement reports directly. Slower to manage and the only option that gives you real control over where a brand appears.
Google Display Ads management
Google Display Ads Management Built on Placement Control
Display inventory is enormous and impressions are cheap, which makes it the easiest place in Google Ads to spend a budget and the hardest place to spend it well. Most of the work is subtractive: deciding where your ads should not appear, and who should not see them.
Run carelessly, a Display campaign reports plenty of activity and produces very little. Run with discipline, it does jobs Search cannot reach.
What Google Display Ads Actually Do
Display places visual and text ads on websites, in mobile apps, in Gmail and across YouTube, shown to people based on who Google believes they are or what they are currently reading. Nobody searched for you. The ad arrives while attention is somewhere else.
That single difference explains most of what makes Display awkward. A person clicking a Search ad has told you what they want. A person clicking a Display ad has been interrupted, and some of them did not mean to click at all. Judging the two channels by the same standard makes Display look like waste even when it is working.
It also changes where the effort goes. In Search you spend your time choosing what to buy. In Display you spend it deciding what to refuse, because the default settings will happily reach a very wide audience across a very wide set of pages.
The honest framing
Display is not a substitute for Search and it rarely produces the same cost per acquisition. What it does is keep a business visible to people who already showed interest, and reach people who have a need they have not yet put into words.
Both are worth paying for. Neither shows up properly if you only read last-click conversions.
The Targeting Layers, Ranked by How Much They Tell You
Display offers several ways to decide who sees an ad, and they are not equal. The ones built from behaviour towards your business carry far more information than the ones built from inference about a person's interests.
-
Your own data
People who visited your site, used your app, or appear on a customer list you upload. The strongest signal available, because it comes from behaviour towards your business rather than an inference about it.
Strongest -
In-market segments
People Google believes are actively researching a category. Useful for prospecting because it implies intent, though the intent belongs to the category rather than to you.
Strong -
Custom segments
Segments you define using terms people search for or sites they browse. Worth building where your category is narrow enough that the standard segments describe it poorly.
Moderate -
Affinity and demographics
Broad interest and life-stage grouping. Better suited to reach than to conversion, and easy to over-trust because the labels sound precise.
Weaker -
Topics and contextual
Placing ads by page subject rather than by person. Sensible where the subject matter itself indicates a need, such as a technical article next to a technical product.
Contextual -
Managed placements
Choosing specific sites, apps or channels yourself. Slow to build, tight in reach, and the only layer where you know exactly where the money went.
Precise
Two practical notes. Affinity and demographic labels sound more precise than they are, and accounts that lean on them tend to buy a lot of indifference. And where optimised targeting or audience expansion is switched on, delivery can reach beyond the segments you chose, so it is worth knowing whether it is running before concluding your targeting was wrong.
Where Display Budgets Leak
Seven recurring sources of waste. None of them announces itself in the campaign summary, and all of them are visible in the placement report if somebody opens it.
| Source | How it happens | Where to look |
|---|---|---|
| Accidental clicks in apps | A banner near a game control or a close button collects taps from people who never intended to click | Placement report filtered to apps, and app category exclusions |
| Low-quality content farms | Pages built to carry advertising rather than to be read, which generate impressions and no interest | Placement report sorted by cost, then reviewed by hand |
| Audience expansion drifting | Delivery reaching beyond the segments you specified, into people with no relationship to the offer | Whether expansion or optimised targeting is on, and what it brought in |
| Wrong geography | Impressions outside the area the business can actually serve, still counted as reach | Location settings, and whether they target presence or interest |
| Uncapped frequency | The same person seeing an ad far more often than is useful, which spends budget without adding reach | Frequency reporting against unique reach |
| Placement in unsuitable context | Ads appearing beside content the brand would not want to be associated with | Content label exclusions and sensitive category settings |
| Weak assets shown widely | Automatically assembled combinations that make little sense, appearing everywhere at once | Asset reporting, and reviewing what combinations are actually served |
The placement report is the single most useful screen in a Display account, and the most neglected. Sorting it by cost and reading the top of the list by hand takes an hour and usually changes where the next month's budget goes.
The Exclusions That Do Most of the Work
Managing Display well is largely a matter of maintaining these lists. They are unglamorous, they are never finished, and they are where most of the recovered budget comes from.
Placement level
- Specific sites and channels that consume budget without returning
- Mobile app categories, and apps individually where needed
- Parked domains and pages that exist to carry advertising
Content level
- Content labels covering material a brand would not sit beside
- Sensitive categories relevant to the business
- Topics that share vocabulary with your category but not its intent
Audience level
- Existing customers, where the campaign is meant to find new ones
- People who already converted, so they stop being paid for twice
- Segments that consistently click and never proceed
Delivery level
- Frequency caps, so reach grows instead of repetition
- Geography set to presence rather than interest where relevant
- Placements at sizes or positions that only collect accidental taps
A caution about going too far. It is possible to exclude so much that a campaign has almost nowhere to serve, at which point delivery collapses and the account looks broken. Exclusions should remove waste you can point to in a report, not everything that feels risky.
Creative for a Channel Nobody Asked to See
A Display ad has to earn attention that was pointed somewhere else, then qualify the person before they click. The second part is the one most briefs forget, and it is why a high click-through rate is not automatically good news here.
Responsive display assets
Google assembles ads from the headlines, descriptions, images and logos you supply, sizing them to fit the space available. That means gaps in what you provide limit where the campaign can appear, and combinations you never previewed will run. Reviewing what is actually being served matters more than approving the assets individually.
Saying the price, or the audience, out loud
Creative that names a price, a market or a constraint gets fewer clicks and better ones. Creative built purely to intrigue gets the opposite. Since clicks teach automated bidding what to pursue, vague creative does not just waste a click, it points the campaign at more people like that.
Uploaded ads where control matters
Where brand presentation is tightly specified, uploaded creative removes the assembly question entirely. The trade is reach, since you only appear where your sizes fit, and maintenance, since every size needs producing and refreshing.
What each asset is doing
ImageStops the scroll, and signals the category before a word is read
HeadlineStates the offer plainly enough to be understood in passing
DescriptionAdds the qualifier that filters out the wrong clicker
LogoCarries recognition, which is most of the value in a reach campaign
Landing pageConfirms the promise for someone with no prior intent
Refreshing before fatigue, not after
Display audiences are finite and see ads repeatedly, so response declines even when nothing has changed. Replacing assets the system consistently avoids, and refreshing the ones carrying delivery, keeps that decline manageable.
Prospecting and Re-engagement Are Different Jobs
Both run on Display inventory and they behave nothing alike. Keeping them in one campaign is how an account reports a respectable average while hiding that only one half is working.
Reaching people who already showed interest
Efficient per conversion, because these people were already close. It is also the part most often over-credited, since it harvests demand created elsewhere in the account. Measured separately it is straightforward to judge. Blended into an average, it flatters everything around it. The mechanics of audience lists and re-engagement sit on Google remarketing.
Reaching people who have never heard of you
Expensive per conversion and responsible for growth. This is where Display earns its keep or does not, and where honest measurement matters most, since the effect frequently appears somewhere other than a last click. Judged on immediate conversions alone, prospecting will always look worse than it is.
Splitting them costs a little data per campaign and buys the ability to answer the only question that matters: is this channel bringing new customers, or recycling ones you already had?
How I Judge Display Without Pretending It Is Search
Display is the channel most often cancelled for the wrong reason. Held to a last-click standard it fails, because a person interrupted while reading something else rarely buys in that moment. Five layers, read together.
-
01
Business outcomes
Revenue, qualified enquiries, and whether total demand moved while Display was running. The blunt test is what happens when it stops.
-
02
Placement and audience quality
Where impressions actually landed and who saw them. A campaign hitting its cost target from app inventory nobody looks at is not succeeding.
-
03
Assisted and view-through effect
Conversions Display contributed to without closing. Worth reading, and worth treating with some scepticism, since a viewed impression is a weak claim on a sale.
-
04
Reach and frequency
Unique people reached against how often each one saw an ad. Rising frequency with flat reach means the budget is buying repetition, not audience.
-
05
Cost, separated by job
Cost per acquisition read separately for prospecting and re-engagement, never as one figure. A blended number tells you almost nothing here.
I will not quote benchmark click-through rates, costs per click or conversion rates for Display. They vary so widely by industry, placement mix and creative that a published figure is worse than no figure, because it invites you to judge your account against something unrelated to it.
Display Ads Compared With Search Ads
The comparison worth making is not which performs better. It is what each one is for, since they are answering different questions about demand.
| Area | Display | Search |
|---|---|---|
| How the ad is triggered | By who a person is or what page they are reading | By what a person typed into Google |
| State of demand | Demand is interrupted or created | Demand already exists and is being captured |
| Main management task | Deciding where ads should not appear | Deciding which queries are worth paying for |
| Typical role | Reach, re-engagement and staying visible between visits | Direct response at the point of intent |
| How to judge it | Assisted and view-through effect alongside direct conversions | Direct conversions and cost per acquisition |
In most accounts, Search should be working before Display is introduced. Capturing demand that already exists is cheaper than creating it, so a business with unclaimed search demand has easier revenue available first. Where that side needs building, it belongs on Google Search Ads.
Where Display Sits Next to Performance Max and YouTube
All three can reach Display inventory. The difference is how much say you have over where the money goes and how clearly you can see it afterwards.
Performance Max reaching Display
Display is one of several channels the system may use in pursuit of a conversion goal, with delivery decisions and reporting both consolidated. Less visibility into individual placements, more reach. How that campaign type is managed sits on Performance Max.
YouTube as its own discipline
Video shares the interruption problem and almost nothing else. Formats, creative demands and the measures worth using are different enough that it deserves separate treatment, which it gets on YouTube Ads.
A useful rule of thumb: if brand safety or placement control is a real requirement, run Display on its own terms. If reach at scale matters more than knowing exactly where the ad appeared, an automated campaign type will get there faster.
When Display Is Worth Running
Worth running when
- You have meaningful site traffic to re-engage, so the strongest audience layer is available
- Search demand is already being captured and you need reach beyond it
- The purchase is considered, so staying visible between visits has value
- Someone will maintain exclusion lists rather than setting them once
- You can judge it on more than last-click conversions
Leave it alone when
- Search demand for what you sell is still unclaimed, since that is cheaper revenue
- The budget is small enough that splitting it would starve both channels
- There is no creative capacity, so assets would run untouched for months
- Conversions cannot be measured, leaving nothing to judge placements against
- The channel will be assessed on cost per acquisition against Search and cancelled
That last point is worth being blunt about. If the plan is to compare Display's cost per acquisition directly with Search and cut whichever loses, Display will lose and the comparison will not have told you anything. Better to agree in advance what it is being asked to do.
How I Manage a Display Account
-
01 Establish
What it is for
Agree the job before the settings: re-engagement, prospecting, or both separated. Then agree how it will be judged, so the conversation in month three is already settled.
-
02 Audit
Where the money went
Placement report by cost, audience layers in use, whether expansion is running, frequency against reach, and what the exclusion lists currently contain.
-
03 Subtract
Remove the waste
Exclusions applied at placement, content, audience and delivery level, based on what the report shows rather than on general caution.
-
04 Sustain
Keep it maintained
Lists reviewed on a schedule, assets refreshed before fatigue rather than after, and reporting that keeps the two jobs separate.
Display Account Examples
Three situations that come up repeatedly, written without figures I cannot show you the workings for.
An account paying for accidental taps
- The problem
- Click volume was high, cost per click was very low, and almost nobody spent more than a moment on the site.
- What it turned out to be
- Most delivery had settled into mobile app inventory, where banners sit close to controls. The clicks were real and the interest was not.
- The work
- App categories excluded, then individual apps as the report identified them, and delivery pushed back towards site placements. Frequency capped so the remaining audience was reached rather than repeated.
- What it led to
- Clicks fell sharply and the account became readable. Cost per click rose, which was the correct direction.
A campaign taking credit for demand it did not create
- The problem
- Display reported a strong cost per conversion, better than Search, which nobody could quite believe.
- What it turned out to be
- Prospecting and re-engagement shared one campaign, and almost all conversions came from people who had already visited through other channels.
- The work
- The two jobs separated into their own campaigns with their own reporting, and existing customers excluded from the prospecting side.
- What it led to
- Prospecting cost per conversion appeared for the first time. The business could then decide what reaching a new customer was worth, rather than assuming it already knew.
A brand appearing beside content it did not want
- The problem
- A stakeholder found the company's ad next to material that was not obviously unsuitable but sat awkwardly with the brand.
- What it turned out to be
- Content label and sensitive category settings were at their defaults, and no topic exclusions had ever been added.
- The work
- Content labels tightened to what the business was actually comfortable with, topic exclusions added for subjects sharing vocabulary with the category, and a monthly placement review agreed.
- What it led to
- Reach narrowed and the internal argument about whether to run Display at all stopped, which was worth more than the reach.
Google Display Ads: Frequently Asked Questions
What are Google Display Ads?
Visual and text ads shown across websites, mobile apps, Gmail and YouTube, delivered to people based on who Google believes they are or what page they are reading. Unlike Search, nobody typed a query first, so the ad interrupts attention rather than answering a request.
Do Display Ads actually work?
They work for particular jobs: staying visible to people who already showed interest, and reaching people with a need they have not yet searched for. They work poorly as a direct substitute for Search, and they fail almost always when nobody maintains the exclusion lists. Whether they are worth running depends more on the job you give them than on the channel itself.
Why do Display campaigns waste so much budget?
Because the default settings reach very widely and the inventory is very large, so a campaign will find somewhere to spend regardless of whether it should. The common leaks are accidental taps in apps, low-quality content pages, delivery drifting beyond the audiences you chose, and uncapped frequency. All of them are visible in the placement report.
What is the placement report and why does it matter?
It shows where your ads actually appeared and what each placement cost. It is the most useful screen in a Display account because it converts a vague sense that something is wrong into a specific list of sites, apps and channels to exclude. Sorting it by cost and reading the top by hand is a genuinely productive hour.
Which Display targeting option is best?
Your own data, where you have it, because it comes from behaviour towards your business rather than an inference about a person's interests. In-market segments come next for prospecting. Affinity and demographic targeting sounds more precise than it is and tends to buy indifference, so I use it for reach rather than for conversions.
Should Display and remarketing run in the same campaign?
No, in almost every case. Re-engagement converts more cheaply because those people were already close to buying, so blending it with prospecting produces an average that flatters the whole campaign and hides whether new customers are being reached at all. Separate campaigns cost a little data and buy a clear answer.
How should Display performance be measured?
Across five layers: business outcomes, placement and audience quality, assisted and view-through effect, reach against frequency, and cost separated by job rather than blended. Holding Display to a last-click standard alone will make it look like waste even when it is contributing, because a person interrupted mid-article rarely buys in that moment.
What is a good click-through rate or cost per click for Display?
I would not give you a number, and I would be wary of anyone who does. These vary so widely by industry, placement mix, creative and audience that a published benchmark invites you to judge your account against something unrelated to it. Your own trend over time is the only comparison worth making.
Are responsive display ads better than uploaded ads?
They reach further, since Google sizes them to whatever space is available, and they need less production work. Uploaded ads give complete control over presentation, which matters where brand appearance is tightly specified, at the cost of reach and ongoing maintenance. Most accounts use responsive ads and review what combinations are actually being served.
Can Display Ads damage brand perception?
It is possible, which is why content label and sensitive category settings deserve attention rather than being left at defaults. The risk is not usually appearing beside clearly unsuitable material; it is appearing somewhere merely incongruous often enough that someone internally notices. Topic and placement exclusions handle most of it.
How much budget does Display need?
Less than the reach available might suggest, and more than a token amount. The practical constraint is whether enough is left after Search is properly funded, since capturing existing demand is cheaper than creating it. A budget too small to sustain both channels is better spent on one.
How often should exclusion lists be reviewed?
On a schedule rather than when something looks wrong, because new placements appear continuously and waste accumulates quietly. The lists are never finished. An account where nothing has been excluded for months is almost certainly paying for something nobody would approve if they saw it.
Want to Know Where Your Display Budget Is Actually Going?
Send me the account with a note on what you sell. I will read the placement report, tell you what you are paying for and what I would exclude first, and say whether Display deserves a budget in your situation at all. Sometimes the answer is that Search should have it instead.