About 99.9% of all banner ads displayed online don’t get clicked on. That’s something to ponder. But before you think: “aha, I suspected they weren’t any good”, let’s put things in context. Firstly, just because they’re not being clicked on, doesn’t mean they’re not being seen. If you’re getting 10 visits a day from your adverts at that 99.9% non-response rate, that means the ad showed 9,990 times without a click. Depending on the positioning on the page, that probably means it was seen several thousand times. And depending on the relevance of the site traffic, it was probably seen several hundred times by prospective customers. Branding is important: prospects are far more likely to click on your Google result or pick up the phone if your name is one they’ve been seeing around the place. So you got a lot more from the advert than your 10 visits.
And a 0.1% clickthrough rate compares with other media anyway. If you take out an ad in a 20,000 circulation magazine and get 20 people to visit your website as a result, you’d think that reasonable. Only an unimpressive proportion of magazines actually get opened, and only a subset of those people will look at the page with your ad. So just as with online banner adverts, if it actually gets seen by a few hundred relevant prospects out of the 20,000 circulation, that’s a decent outcome. Banners are competitive, therefore.
The issue with banners is how you pay for them. Some sites will charge you a fixed rate, just like magazines do, to run your advert. Now, with controlled circulation magazines, at least you know how many copies are being mailed out in advance. With banner adverts online, you have no guarantee how many people are going to visit the site. So a fixed-rate model, just to run a banner advert for a set period with no guarantees, is really quite unacceptable nowadays. Unless you’ve had consistently good results with a particular site in the past, measured at your end, then just say no to this.
A “cost per thousand impressions” payment model is better, and closer to magazine advertising. Here, you pay for your ad to be seen a certain number of times, much as with magazines (where you pay to be included in the known number of copies being mailed out). This route is the minimum which should be acceptable, although you must set the ad up so that you’re measuring the number of times it’s seen, not relying on the website owner’s data.
An alternative, and one which was dropped by magazines many years ago, is what the publishing industry used to call “payment by results”. For banner advertising, this means paying just for the clickthroughs to your website, regardless of how many times the advert gets seen. Unless you’re advertising primarily for branding purposes, this is what you should be looking for.
“Pay per click” advertising is offered by all the big social sites such as LinkedIn and Facebook, as well as by the big third party brokers (a field dominated by Google AdWords). You can use Google AdWords to book “pay per click” banners on many important websites, alongside thousands of smaller (but relevant) ones which you wouldn’t normally have the time to deal with. If you’re still just doling out advertising budget willy-nilly to websites which insist on fixed rates and don’t charge you solely for what they deliver, perhaps 2014 is the time to start spending your money more wisely.