Monday, December 08, 2014

In Praise Of Banner Ads

There was a bit of a hoo-ha last week, when Google announced that 56.1% of all paid-for display ads never actually appeared on anyone's screen.

"Google admits that advertisers wasted their money on more than half of internet ads," was the view of a typical commentator

But are the attacks on banner ads justified?

Yes, it's true that click-through rates are low.
Perhaps you've seen a list that's been going around the interwebs - '5 Horrifying Stats About Online Display Advertising'.
It goes like this: the average banner ad has a 0.1% clickthrough rate, therefore…

  • You are more likely to be dealt a full house in a poker hand than click on a banner ad. (Source: Solve Media)
  • The average person is served over 1,700 banner ads per month. Do you remember any? (Source: comScore)
  •  You are more likely to give birth to twins than click a banner ad. (Source: Solve Media)
  • About 50% of clicks on mobile ads are accidental. (Source: GoldSpot Media)
  •  You are more likely to survive a plane crash than click on a banner ad. (Source: Solve Media)

    Is there anything good about display advertising?

    Yes. First off, ads appear next to the content you're already consuming and are interested in. That's handy. Also, retargeting can be used to hit people who have already visited a website; that means you're targeting a pretty hot prospect.

    Most of all, they work. Their results are highly measurable. So we know they work.

    Click through rates are indeed low on banner ads, and have been plummeting for years. The very first ones, like the 1994 ad shown above, which was for AT&T and is thought to be the very first banner ad ever served, had click-through rates as high as 10%.

    The average is a hundredth of that now.

    But hey, we can’t click on TV ads or radio ads or outdoor ads either.

    And as for that stat about 56% of online ads not being viewed... surely it's the same with print.

    What percentage of the ads do you see, when you read a newspaper or magazine? Chances are that you don't flick through every page of the newspaper, therefore the ads on those pages never appeared on those screens that you commonly call your eyeballs.

    Same as with banner ads.


    adhut said...

    What has been exposed here? The uselessness of banner ads or the uselessness of print, radio, TV or outdoor?

    Actually neither. Make your more ads as interesting than the media they appear in has always been a good rule to follow.

    Anonymous said...


    Banner ads therefore (by scamp's rationale - which is a perfectly fair one one) are reduced to just another display medium... which is what they always were to be honest.

    So half of one's advertising is still wasted... and we still don't know for sure which half.

    Once again we are reminded that just because something can be very accurately measured, that doesn't make it meaningful.

    ....Like I said - Interesting, but in retrospect, also hardly surprising.

    Scamp said...

    Very true - we still don't know for sure. And yet, it is considerably more possible to measure whether someone bought something after seeing a banner, than after seeing a print ad.

    Anonymous said...

    I'll just copy paste this here Simon, since you've been missing the point of that Google article entirely:

    "As I'm sure you know, all forms of advertising are subject to waste. This is because not everyone notices every ad. This is just the nature of advertising. Apologists for online advertising try to excuse these problems as just another example of ad waste. It is not.

    The waste in online ads is of another magnitude. First, criminals are stealing your money. Then, unviewable ad placements are siphoning more of your money. And after all that, what's left is still subject to the normal waste of advertising."

    Anonymous said...

    Is that you Bob?

    Anonymous said...

    No it's not Bob, but I copy pasted it from his blog.

    Anonymous said...

    "People are criticising x, but y is just as bad"

    -anonymous, 'in praise of x'

    Anonymous said...

    All media has high wastage. Digital Display's problem is that it's uselessness can be measured. I walk past an APN Outdoor digital display every day enroute to work, as do thousands of other commuting folk. By evening, I'd bet much less than 56% of us can recall a single Ad we saw that morning. The problem here is that digital media strategy and analytics is so far behind other markets that high CPM premium buys without any sort of conversion measurements are just the norm for a lot of clients and media buying shops.

    Mike Fletcher said...

    The online ads weren't 'not viewed' as you state, they 'never actually appeared on anyone's screen' as Google confirm.

    There's a massive difference between an ad not being seen and an ad that doesn't even appear when a business has paid for it to do so.

    Scamp said...

    I don't think there is a difference - I guess that's the whole point I'm trying to make. An advertiser pays for their ad to appear in either The Guardian newspaper or on In both cases, the ad is there, waiting to be seen. If the consumer doesn't turn the right page, they never see the print ad. If they don't scroll through to the right bit of the website, they don't see the banner. Same same.

    Anonymous said...

    Simon, it's not that they don't see the ad because they don't scroll down. It's because the ad runs in a 1x1 pixel sized square. It literally cannot be seen, even if you happen to scroll to that part of the site.

    The 1x1 pixel size is just one of the many tricks by the way. There's also a way to stack ads on top of each other, so only the top one is visible but they actually bill you for the ones that are on the bottom and can't be seen.

    Scamp said...

    You're talking about fraud. Which certainly exists, of course. But I don't think the Google article is saying that 44% of banner ads are fraudulent.

    Anonymous said...

    You're right Simon, but do you think you can exclude the banners which count as "viewable" from fraud and thus messing with that % number even more?

    After all, that 1x1 pixel sized banner was "viewable", by all accounts and measures.