Display advertising is the practice of placing visual advertisements, such as static banners, animated creatives, rich media and video, in designated ad placements on websites, apps and connected screens. In digital advertising, it complements search engine marketing: search ads answer an expressed need, while display ads reach people before they search, shaping awareness and consideration across the open web.
Display campaigns are bought either directly from publishers or, far more often today, through programmatic advertising, where software buys individual impressions in milliseconds. Pricing is usually based on CPM, the cost per thousand impressions, although CPC and CPA models are also common in marketing practice.
How display advertising reaches an audience
Display advertising works by matching an ad creative with an available impression at the moment a page or app screen loads. The publisher’s page contains an ad slot defined by size and position. When a visitor arrives, the slot sends an ad request, and either a pre-arranged campaign or an automated auction decides which advertiser fills it.
In programmatic buying, the request travels from a supply-side platform (SSP) to one or more ad exchanges, where demand-side platforms (DSPs) evaluate it on behalf of advertisers. Each DSP estimates the value of that impression from contextual signals, audience data and campaign goals, then submits a bid. The entire process, known as real-time bidding, typically completes in about 100 to 200 milliseconds.
The table below shows the main parties and their roles in a programmatic display transaction.
| Participant | Role | Main goal |
|---|---|---|
| Advertiser | Funds the campaign and supplies creatives | Reach and conversions at target cost |
| DSP | Bids on impressions for advertisers | Buy the right impressions efficiently |
| Ad exchange | Runs the auction between buyers and sellers | Clear transactions at market price |
| SSP | Offers publisher inventory to buyers | Maximize publisher revenue |
| Publisher | Owns the site or app and ad slots | Monetize audience attention |
| Ad server | Delivers and tracks the creative | Accurate delivery and reporting |
Display ad formats and their uses
Display formats range from simple static images to interactive experiences, and the choice affects production cost, page performance and engagement. The IAB New Ad Portfolio and older IAB standard sizes define common dimensions so that one creative can run across many publishers without custom work.
Responsive display ads, which assemble headlines, images and logos automatically for each placement, have become standard on large networks because they reduce creative production effort while fitting thousands of slot shapes.
| Format | Typical size or behavior | Best suited for |
|---|---|---|
| Medium rectangle | 300 x 250 pixels | Broad reach, in-content placements |
| Leaderboard | 728 x 90 pixels | Top-of-page desktop visibility |
| Mobile banner | 320 x 50 pixels | Low-cost mobile reach |
| Rich media | Expands or reacts to interaction | Engagement and product demos |
| Native display | Matches surrounding content style | Less intrusive consideration ads |
| Outstream video | Plays inside article text | Video reach without video sites |
Targeting and pricing models
Display targeting defines who sees an ad and in what context. Contextual targeting matches ads to page content, audience targeting uses demographic, interest or intent segments, and retargeting shows ads to people who previously visited the advertiser’s site. With the decline of third-party cookies and stricter privacy rules such as the GDPR, contextual methods and first-party data have regained importance.
Pricing models align cost with different objectives, and choosing the wrong one distorts campaign optimization. The following table compares the main models used in display buying.
| Model | Advertiser pays for | Typical objective | Main risk |
|---|---|---|---|
| CPM | Every thousand impressions | Awareness and reach | Paying for unseen ads |
| vCPM | Thousand viewable impressions | Quality reach | Higher unit price |
| CPC | Each click on the ad | Site traffic | Accidental or fraudulent clicks |
| CPA | Each completed action | Sales and leads | Limited inventory access |
Display click-through rates are low, commonly around 0.05 to 0.1 percent for standard banners, although they vary widely by format, placement and audience. For this reason, experienced planners judge display mainly on reach, viewability and incremental lift, not on clicks alone.
Display advertising compared with search and social ads
Display advertising differs from search engine marketing mainly in user intent. Search ads appear when someone types a query, so intent is explicit and conversion rates are high. Display ads interrupt browsing, which makes them stronger for awareness and weaker for immediate response. Social advertising on platforms such as Meta Ads sits between the two, combining rich first-party audience data with feed-based formats.
| Channel | User state | Strength | Typical KPI |
|---|---|---|---|
| Display advertising | Browsing content | Scale and visual impact | Reach, viewability, lift |
| Search advertising | Actively searching | High purchase intent | Conversions, CPA |
| Social advertising | Scrolling a feed | Detailed audience data | Engagement, CPA |
| Video advertising | Watching content | Storytelling and emotion | View rate, recall |
Limitations, risks and measurement pitfalls
Display advertising carries risks that do not affect search to the same degree. Banner blindness reduces attention, ad blockers remove inventory, and ad fraud such as bot traffic and domain spoofing inflates impressions. Standards such as ads.txt and sellers.json help buyers verify authorized sellers, while viewability measurement checks whether ads could actually be seen.
The industry viewability threshold defined by the MRC is at least 50 percent of pixels in view for one continuous second for display ads. Many impressions fail it, so buying only viewable inventory is a common safeguard. The table below lists typical problems and remedies.
| Problem | Likely cause | Remedy |
|---|---|---|
| High impressions, no results | Low viewability or fraud | Buy viewable inventory, use verification |
| Ads next to unsafe content | Weak brand safety controls | Apply exclusion lists and context filters |
| Inflated conversion credit | Last-view attribution | Run incrementality tests |
| Users annoyed by repeated ads | No frequency cap | Limit exposures per user per day |
| Slow publisher pages | Heavy creatives | Compress files and limit animation |
The most serious measurement mistake is crediting display with conversions that would have happened anyway. View-through attribution overstates impact, so mature teams validate results with holdout or geo experiments.
A plain-language summary
Display advertising is the digital version of a billboard or a magazine ad. When you read a news site or use a free app, the pictures and short videos in boxes around the content are display ads.
Companies pay to put them there so that people remember their name or product. Most of these spaces are sold automatically in tiny auctions that happen while a page loads. The main purpose is to be seen and remembered, not necessarily to be clicked right away.
An example from practice
A regional furniture retailer wants to promote a new sofa collection before the holiday season. Its marketing team runs a six-week programmatic display campaign through a DSP with a budget of 40,000 dollars, targeting home and interior design content within 50 kilometers of its stores, plus retargeting of recent site visitors. Creatives include 300 x 250 and 320 x 50 banners and a short outstream video.
The team buys only viewable impressions, sets a frequency cap of four exposures per person per day, and keeps 10 percent of the target audience as an unexposed holdout group. After the campaign, store visits and online orders in the exposed group are about 12 percent higher than in the holdout, which gives the retailer a credible estimate of incremental sales instead of relying on click counts.