View Single Post
Old 08-14-2007, 07:29 PM   #1
the_logos
Legend
 
the_logos's Avatar
 
Join Date: Sep 2002
Location: Mill Valley, California
Posts: 2,305
the_logos will become famous soon enough
Reaching out beyond text MUDs

There's been some good discussion lately about reaching out beyond text MUDs. I don't know how many of you actively try to spread the message beyond existing text MUD users but I thought I'd share the specific results of an experiment we just ran.

We want to start paying an ad agency per registration they deliver to us. To be clear though, by registration we don't mean someone who just creates a character. In our games, registration is when you're asked for your rl details, which doesn't happen until you've finished the newbie intro for the game. So, a player has to stick it out through the newbie intro before we'll pay this company for that player.

In order to set a price per registration what we did was run a small test where they'd deliver 11224 click-throughs to us (there's a reason we used 11224 but it's not interesting or relevant for this discussion). I believe it took about 1.4 million impressions to deliver those clicks. (For reference, the actual creatives we used can be found .)

Each of those clicks went to the and from there to opening our Nexus client, then to creating a character, then into the game and finally registering. The numbers worked out like this:

Those 11224 clicks cost us 5 cents/click, which means we paid $561 for 51 registered players, or about $10/player. However, the source site mattered greatly. For instance, on one set of sites we got closer to about $5/player while on others it was more like $30/player. The small sample size of registered players probably guarantees large statistical errors here though.

We don't have any idea yet if those registered players are actually worth $10 each of course and won't truly know for over another year (as we have to watch how people spend first).

The first big problem I see here is our portal. A ~20% rate to just click on one of the 'Play Now' buttons is terrible and we're going to completely redo our portal to try and improve this. Shame, as I rather like our portal but the numbers are pretty unequivocal.

60% of players completed the character creation process which is fairly good I'd say, so I'm not too worried about this aspect of things. I'm guessing it's high because the creation process is graphically-driven in Nexus, with character portraits and that sort of thing to select from.

Then things get ugly again. 51 out of 1297 players is about 3.9%. That means 96.1% of the players who finished character creation dropped out soon after being put straight into a mainly text environment (our Nexus client puts at least somewhat attractive graphics around the text output area but it's still mainly a text output area, of course).

It's very hard to know what the issues are in the last part. The fact that players are suddenly in a text environment when they were previously on a graphically-oriented webpage or in graphically-driven character creation is almost certainly a big part of it. Who knows what else plays into it though. There are so many factors that could affect this, from the style of prompt a MUD uses to the colors (or lack thereof) it chooses to the length of room descriptions, down to whether it uses UK or American spellings (colour vs. color), etc etc etc. In theory we could change one of these at a time and run tests to optimize but that's not really economical sadly.

What we're considering is possibly having illustrations done that give the general 'feel' of the general area you're in when you start playing, and display those in Nexus. We'd then display illustrations less and less as the person goes on in the game until finally (probably once you're done with the newbie introduction) they're not getting any at all. The idea here is to kind of slowly shepherd someone from a mainly graphical environment (character creation) to a fully text environment and minimize how 'jarring' the experience of going from one to another is. I have no idea how effective this will be (or even, at this moment, of whether we'll go ahead with it or not) but it's what's come up in discussion after seeing these numbers.

Anyway, just thought I'd share.
the_logos is offline   Reply With Quote