I’ve been casting my geeky eye over the google analytics data for the gratuitous tank battles website. Here are some assorted stats…

Looking at data for the last 60 days, the site had:

Visits: 67,440

Pages per visit: 1.49

Bounce Rate: 74.78%

TBH, only one of those stats is worth caring about (the top one), and only then, marginally. Bounce rate and Pages per visit are horrendously skewed by the content of those pages. If you pack them with text and video, people will get their fill of data with just one page. These are not stats worth worrying about. The visits stats is marginally more helpful, but it massively depends where they come from, obviously. I’ve sponsored a few small flash games which brings in a TON of traffic, but most of it is pretty ‘low quality’, kids without credit cards looking for more free games. The sheer volume means it can be worth it though. I find it more helpful to concentrate just on highly engaged visitors, such as those spending more than 60 seconds on the site. That gives me:

Visits: 4,723

Which are the only site visitors I really care about. This means I need to know where they came from. In this case the two big easily identified source areas seem to be google adwords and google organic search, which are roughly equal. The problem is, this isn’t showing me my flash sponsorship traffic correctly, so I need to do some analytics cleverness to detect when the ?ref= parameter is passed which tells me which flash game sent me the click. when i look for those clicks I find they supplied…

Visits: 2,977

Which is clearly the lions share of those above. Of course, all this means is those visitors spent time on the site, unfortunately I don’t have any easy way to tell that they are the same people buying the game, especially if they drift off and buy it on steam, or next month. However, I am quite motivated by the long tail effect here. There are hits coming in from long forgotten sponsorship deals from ages ago. That doesn’t happen with banner adverts which are obviously immediate. This can be a pain in the neck, because banner ads can be scheduled and also ramped up and cut back to fit your budget, whereas free game sponsorship is a bit of an all-or-nothing, no-idea-when sort of deal.

Even so, I think I’ll keep experimenting with it. Unlike my foray into stumbleupon and facebook ads, I think this may actually have a reasonable (on a good day) Return-on-investment

2 Responses to “Gratuitous Tank Battles website stats”

  1. Werit says:

    Very neat info, thanks for sharing!

    For analytic’s, I really like http://getclicky.com/ provides a lot of tools and a decent api.

  2. Andrej says:

    Thanks for sharing this insight, this is something that wouldn’t have occurred to me on my own. I knew making a promo flash game worked for a few developers to drive traffic and revenue of the main game, but this is a really elegant solution.

    The benefit of helping another developer working on the flash game is an added value to this approach.