An old topic revisited today, but people are still putting adverts on third-party websites and then not measuring the clickthroughs. Then they ask me: “Can you explain why the publisher of Blue Widgets Online reckons my banner ad got 593 clicks last month, whereas Google Analytics says the site sent me just 6 visits?” What can I say? I’m usually diplomatic, and say that the publisher may have made a mistake.
In fact, what some publishers do is to route anyone who clicks on your advert to an intermediate page, which then invisibly bounces them through to your site. Then they count the views of this intermediate page as representing the number of clicks on the advert. This makes sense, until you remember that lots of the traffic following this path won’t be real people. Indeed, most of it won’t be. You wouldn’t believe how much of the traffic flowing around the web is automated ‘bots’ and ‘crawlers’. So the counter at the publisher’s end ticks over, but Google Analytics at your end doesn’t (because Google Analytics knows who’s real and who isn’t). The publisher gives you a number – as measured by their system – which technically equates to the number of ‘clicks’ on your advert, without admitting that most of these ‘clicks’ aren’t real people.
So let’s just measure our adverts on third-party websites properly, shall we? If you’ve got Google Analytics running, give the publisher a separate, tagged URL for each advert, so you can see how many real visitors are coming in from each advert type. Then set things up so that you can see how many of them are any good. There are two steps here:
1. Tag the adverts. Don’t tell the publisher to send traffic to www.bluewidgetcompany.co.uk, but instead use Google’s URL Builder to add some tagging to give you better data in Google Analytics. You might set the website’s name as the source, “banner ad” as the medium, and the product, ad size or page location as the campaign. There are many good guides online: here’s one.
2. Start measuring quality visitors. Google Analytics can measure anything. But a simple start would be to measure the visitors who came from the countries you want, and who stuck around on your site. I explained how to do this here as part of a reporting exercise last year.
Now, instead of accepting a meaningless figure on clickthroughs provided by a publisher, you’ll be able to say: “I received 37 visits from the banner ads on your site, and 11 were potentially useful ones”. Isn’t that a better figure on which to base a decision to re-book advertising – and how much to pay for it?