Banner ads have a reputation problem that is mostly a measurement problem. Cross-industry display click-through rates sit around 0.46%, and standard banners bought on the open programmatic exchange run closer to 0.05–0.1% (WordStream, Display & Video Ads Benchmarks, retrieved 2026-10-02). Read as a direct-response channel, those numbers look like a catastrophe. Read as what they are, they are ordinary: an awareness format where 995 of every 1,000 viewers were never going to click.
That distinction decides whether banner ads work for you. Bought to generate clicks, they will disappoint. Bought to make a name familiar before a buyer searches for it, they do a job nothing else does as cheaply. This covers the sizes worth buying, what the creative has to achieve in a single glance, where to place them, and how to judge the result honestly. Display sits near the end of the sequence in the ways to market your business and attract local customers, and it earns its place only once the direct-response basics are catching demand.
Key Takeaways
- Banner ads are an awareness format. Judge them on whether brand searches and direct enquiries rise, not on clicks.
- Four sizes do almost all the work: 300×250, 728×90, 160×600 and 320×50. Buying the full IAB catalogue wastes design time for marginal inventory.
- A banner has roughly one second. One idea, one readable line, a visible logo, and a button. Anything more is invisible at that size.
- Published display benchmarks are shakier than they look, because much of the circulating data traces to datasets last refreshed in 2018 or earlier, so your own baseline matters more than any industry average.
- Rich media and native formats are reported to earn roughly 4x and 2.5x the click-through rate of standard banners, which is usually a better lever than another round of creative tweaks.
Metrobi drivers are rated 4.97/5
Trusted by local businesses for:
- Background-checked professionals
- Specialized in business deliveries
- Same drivers for consistency
- 4.97/5 average delivery rating
What a banner ad has to do in one second
A banner ad has to deliver one recognisable idea at a glance, from the corner of a page the reader is not looking at, to someone who did not ask to see it.
That constraint is the whole brief, and it explains why most banner ads fail. They are designed as miniature posters (three lines of copy, a feature list, two product shots) on the assumption of a viewer who stops and reads. Nobody stops and reads. The viewer is mid-scroll, their attention is on the article, and your ad occupies a space they have been trained for twenty years to ignore.
So the test for any banner creative is brutally simple: cover it after one second and see whether you can state who it was for and what it offered. If not, it is too complicated, regardless of how good it looks at full size on your monitor.
The banner ad sizes worth buying
A handful of standard sizes account for the overwhelming majority of available inventory. Build these and you can stop.
| Size (px) | Name | Where it runs | Priority |
|---|---|---|---|
| 300×250 | Medium rectangle | In-article and sidebar, desktop and mobile | Essential |
| 728×90 | Leaderboard | Top of page, desktop | Essential |
| 160×600 | Wide skyscraper | Desktop sidebar, stays in view while scrolling | Important |
| 320×50 | Mobile banner | Anchored bottom of mobile screens | Essential |
| 336×280 | Large rectangle | In-article, slightly better performing than 300×250 | Useful |
| 320×100 | Large mobile banner | Mobile, double height | Useful |
| 970×250 | Billboard | Premium top-of-page desktop | Optional |
| 468×60 | Full banner | Legacy format, declining inventory | Skip |
| 234×60 | Half banner | Legacy format | Skip |
| 120×240 | Vertical banner | Legacy format | Skip |
The four marked essential or important will reach nearly all the inventory you can realistically buy. The 300×250 is the single most useful because it runs on both desktop and mobile and sits inside article content rather than at the page edges. The 320×50 matters more than its size suggests, because most of your audience is on a phone.
One design consequence: your creative has to survive being both a wide strip and a tall column. Design the 300×250 first and make it good, then adapt. Designing for the leaderboard first produces a layout that cannot be rebuilt vertically.
What click-through rate is realistic
Set expectations from the data, then build your own baseline, because the published benchmarks are less solid than they appear.
| Format | Reported average CTR |
|---|---|
| Standard display, cross-industry | ~0.46% |
| Standard banners, open programmatic | ~0.05–0.1% |
| High-impact display | ~0.26% |
| Native display | ~0.28% |
| Video | ~0.09% |
| Real estate (highest vertical) | ~1.08% |
| Technology | ~0.39% |
Figures compiled from WordStream’s display and video benchmarks and aiDigital’s display CTR analysis, retrieved 2026-10-02.
Now the caveat that most guides omit. There is no reliable current industry benchmark for display click-through rates. Much of the data in general circulation traces back to datasets last updated in 2018 or earlier, and gets re-quoted each year with a new date attached. Treat the numbers above as a rough order of magnitude for setting expectations, not as a target to be measured against.
What this means practically: your first campaign’s job is to establish your own baseline. Run it, record the rate, and compare every subsequent campaign to that figure rather than to an industry average assembled from a different mix of advertisers, placements and years. Above 0.5% is generally considered solid and above 1% exceptional, but your own previous month is a better comparison than either.
The format choice is also a bigger lever than the creative. Rich media and native formats are reported to earn roughly 4x and 2.5x the click-through rate of standard banners respectively, so moving format will typically outperform a fourth revision of the same static image.
Design rules for a banner that works at small sizes
Four rules cover most of it.
- One idea, nothing else: not the product range, not three benefits. A single offer or a single claim. If you cannot decide which, that is the real problem and more design will not solve it.
- Readable at actual size: set the type, then view the banner at 100% on a phone. Most banner copy is set at a size that is legible only in the design file. Roughly 24px minimum for the headline in a 300×250.
- Logo visible, always: since most viewers will never click, the brand impression is the return. An ad that gets ignored but leaves a name behind has done its job; an unbranded one has not.
- A button, even though it is barely clicked: the button signals that the thing is clickable and gives the eye somewhere to finish. High contrast, short verb, no cleverness.
If you are short of starting points, galleries of AI ad creative examples are a quick way to see how much a single-idea layout can carry at small sizes. Real photography beats stock and illustration here, though, and for a local business it is also free. A photograph of your actual van, your 5am kitchen, or the arrangement that went out on Saturday carries information a stock image cannot and cannot be run by a competitor. That is the same argument that runs through the strongest advertisement examples to borrow from: the transferable moves are structural and cheap, not expensive.
Writing the one line of banner copy
You have room for about six words that will actually be read. Spend them on something checkable.
Superlatives are the default and the weakest option. “Best bakery in the city” is unverifiable and gets discounted automatically. “Baked at 4am, delivered by 7” is specific, falsifiable, and tells a reader something they can act on. The same logic favours a concrete constraint over a vague promise: “12-mile delivery radius” does more work than “fast local delivery.”
Write the line around whatever you offer that the alternatives do not, which means knowing what that is. If you have not pinned it down, reading the market structure you compete in is the exercise that produces it. In a market of many similar sellers, the specific difference is the only thing worth putting in six words.
Where to place banner ads
Placement quality affects results more than creative quality, and it is the part small advertisers most often delegate entirely to an ad network.
- In-article, above the fold is the strongest general placement. The reader is engaged with the page and the ad sits in the reading path rather than at the periphery.
- Anchored mobile placements stay on screen as the user scrolls, which produces high viewability. Use sparingly: they are also the most irritating, and irritation attaches to your brand.
- Sidebar skyscrapers work when they remain in view during scrolling and are largely wasted when they scroll away in the first screen.
- Local and niche publishers usually outperform broad networks for a local business. A city food blog or a neighbourhood news site reaches a relevant audience at low cost, and you can often buy directly, which removes the exchange’s cut.
Two things to set before you spend. Cap frequency, because the same person seeing your banner twenty times produces annoyance rather than familiarity. And exclude the placements you do not want, because most networks default to including app inventory and made-for-advertising sites where impressions are cheap and worthless.
If you run your own display campaigns you will also need the consent layer in place, since serving tracked advertising to visitors in most jurisdictions requires a compliant cookie banner collecting consent before tracking begins. That is a prerequisite rather than an optimisation.
Judging banner ads without fooling yourself
Because clicks are a poor measure of an awareness format, you need two or three indirect signals instead.
Watch whether branded searches rise, meaning people searching your business name rather than a category term, during and after a display campaign. Watch direct enquiries and phone calls over the same window. Watch whether your other channels get cheaper, since familiarity makes search ads and emails convert better. Those are the mechanisms by which display actually pays, and none of them appears in a click report.
Set this up before you buy anything, or you will end up making the standard mistake: cancelling a campaign that was working because the only number you had was the click-through rate. The six numbers worth tracking and how to connect a sale back to its cause are covered in marketing analytics for small business.
Be clear-eyed about the economics too. Display is reported to return around $1.35 per $1 spent, the lowest of the major channels. That is a real constraint, and it is why display belongs after the free local foundations and after email, not instead of them.
When banner ads are the wrong choice
Display is the wrong first channel for most small businesses, and recognising that saves money.
If you have no email capture, no complete Google Business Profile, and no review habit, display spending will generate awareness that leaks straight back out because nothing is set up to catch it. If you need orders this week, search ads reach people who have already described what they want and will do it faster. If your budget is under roughly $200 a month, display will spread too thin to build the frequency awareness requires, so put it into the free foundations instead.
And if what you actually need is to answer the questions standing between a buyer and a first order, that is a job for owned content rather than bought impressions. Content marketing for small business is slower but keeps working after you stop paying, which is the opposite trade-off to display.
Frequently asked questions
What is a good click-through rate for banner ads?
Around 0.46% is the commonly cited cross-industry average, with above 0.5% considered solid and above 1% exceptional. Treat these loosely: much of the circulating benchmark data traces to datasets last refreshed in 2018 or earlier, so your own first campaign’s rate is a more meaningful comparison than any published figure.
What are the standard banner ad sizes?
The ones worth building are 300×250 (medium rectangle), 728×90 (leaderboard), 160×600 (wide skyscraper) and 320×50 (mobile banner), with 336×280 and 320×100 as useful additions. Legacy formats like 468×60, 234×60 and 120×240 have declining inventory and can be skipped.
Do banner ads still work in 2026?
Yes, as an awareness format. They build familiarity that makes your other channels convert better, which is why they should be judged on branded search volume and direct enquiries rather than clicks. As a direct-response channel they are weak, with reported returns around $1.35 per $1.
How much should a small business spend on banner ads?
Below roughly $200 a month, display spreads too thin to build the repeated exposure awareness requires, so the money is usually better spent on the free local foundations or on search ads. Above that, add it as a single channel, cap frequency, and hold it for at least three months before judging.
Why do my banner ads get impressions but no clicks?
Because that is how the format behaves: roughly 995 of every 1,000 viewers will not click. If clicks are the goal, switch to search advertising or move to a rich media or native format, both reported to earn several times the click-through rate of a standard banner. If awareness is the goal, the impressions were the product.
The honest summary
Banner ads work, but not in the way they are usually sold. They are a familiarity purchase: you are paying so that your name is already known when someone searches your category next month, which makes everything else you do cheaper.
Build the four sizes that matter, put one checkable claim and a real photograph in each, buy local and niche placements rather than broad networks, cap frequency, and measure branded search instead of clicks. Then judge it against your own baseline rather than a benchmark from 2018.
