Fun with conversion rates

[princec]

Check the Ultratron conversion ratio out*:

Version Sales Downloads CR
1.0 119 6108 1.95%
1.1 8 204 3.92%

Remember I adjusted the game timeout configs to pick the ones that were 5x better than the worst performing ones? And also tweaked the nag so the game timed out sooner? Well there's the result: CR doubled. Early days yet of course.

Cas :)

* excluding Arcadetown (logging is different for Arcadetown so can't draw comparisons)

[Nexic]

Thats really great! However the second, 4% CR is really a very small sample size. I don't know how long its been since 1.1, but it is highly possible that a someone who dowloaded 1.0, and bought the full version, just after you release 1.1 (but didnt get it).

For example my monthly conversion report from a portal usually come in pretty low (under 0.5%), but this month I got one saying 3%. It turns out that they took it off their site near the start of this month, but people who had downloaded the previous month bought the game, and that pushed the CR right up.

Just out of interest, are your 127 sales all direct? And what total time period?

[princec]

All direct, and since 6th June 2005 (2 months on sale!)
I can't detect people playing 1.0, upgrading to 1.1, and then buying it, but by and large that's exceedingly rare I think. So the CR for 1.1 is looking good at this stage given that no other factors have really changed. The logs are cunning in that they're pretty accurate at telling me an accurate CR from a particular version of the game though. These figures are far more accurate than simple downloads/sales * 100.

Cas :)

[Nexic]

63.5 direct sales per month, and a 2%+ CR. Damn I'm seriously jealous :)

Other than ArcadeTown are no other portals interested in running your game, or do you not want them there?

(I really don't want this comment to start a huge 'make cas sell on portals' thread. Just want to know what the portals stance might be on this kind of game)

[princec]

Currently none of the big portals are interested (tried, got told "no"). Have to strike a deal with Reflexive but I'm waiting until I've got that CR good and tuned :)

Might try an ad campaign on Apple.com if I can afford it. Any other suggestions for an ad campaign that's focused on profits rather than gaining "new customers"?

Cas :)

[Robert Cummings]

Thanks for sharing cas :)

So you made the demo time out before an hour was up, or was it some other factor? I didn't clearly understand what you did. But whatever it was, seems to look good!

[princec]

Demo configures itself at install to choose a random number of max. games and random max play time (nowhere near an hour either! More like 10 minutes).

Oh bugger, there's another sale :) That's screwed the figures up again.

Cas :)

[Anthony Flack]

Hooray! Except I think that 8 sales is too small a sample size to attach any statistical significance to.

[Robert Cummings]

Perhaps his grandad was first over the top. But we're all grateful for it :)

[princec]

Yeah, 8 is early days yet... just watching the sales perk up though made me all enthusiastic :)

Bad news is that more cunning logging shows an increase in failure rate. Boo!

Interesting story: my great grandad was described as being killed by "a bomb" early in WWI, which, as bombs dropped from planes were a brand new way to kill people back then, made him one of the first people to be killed in this way.

Cas :)

[Robert Cummings]

I meant great grandad too, but didn't edit out of laziness. Suprising coincidence or weird spooky twilight zone moment? well, you're officially destined to be one of the first, come what may...