Download Size is Irrelevant
April 6th, 2006 at 4:18 pm
I’ve whipped up a file downloader that helps me track how often the Jeweltopia demo is downloaded, how fast it was downloaded, and whether or not the download completed, with the capacity to determine between a canceled download and a timed-out download. After a few weeks with targetted traffic (casual gamers), I’ve come to some stunning conclusions. First, here are the raw stats:
- Users with broadband: 89%
- Users with dial-up: 11%
- Broadband Success Rate: 94%
- Dialup Success Rate: 38%
- Overall Success Rate: 88%
Now, it’s important that I point out that the Jeweltopia demo is a mere 4.72MB in size — certainly below the age-old maxim of 5MB — so I’m not scaring away dial-up users with a huge download. At 4 KB/sec (the average dial-up speed), the demo should be expected to take approximately 20 minutes to download — a reasonable investment for the typical dial-up user. It should be no surprise, then, when I tell you that 63% of the canceled dial-up downloads were transferring at under 1 KB/sec and lasted for at least 30 minutes. Even folks on bad connections were certainly willing to at least give it a shot with a 30 minute investment.
However, the failure rate for dial-up users is shocking, to say the least. Almost two thirds (62%) of dial-up users canceled their download. Compare that to only 1 out of every 20 broadband users canceling their downloads, most within the first minute. It certainly appears as though dial-up users are least likely to complete a download, despite being more patient than their high-speed bretheren. No matter the connection, however, the vast majority of users who canceled their download did so at under 15% completion, or about 708 KB. Less than 1% of canceled downloads were canceled after the user had downloaded 15% (708 KB) or more of the file.
While a 1 MB download would probably capture the vast majority of users, that’s far too small to effectively compete in this industry. However, this data indicates a strong propensity for users to accept large downloads and sub-700 KB web-games. How big can we go? I can’t yet tell, because I have not yet tried other file sizes and don’t know the psychological ramifications of such things, nor have I been able to witness the upper threshold of broadband downloaders. However, if we target a larger file-size, we can probably assume we’re only further alienating dial-up users. As such, let us temporarily remove them from our considerations.
The average broadband user has a 200 KB/sec download capacity, or approximately a 1.6 Mbit connection. If we project the dial-up user fail rate onto broadband users — 700 KB @ 1 KB/sec, or approximately 11 minutes of download time — we find ourselves with an intriguing result: 132 MB. Does this mean we can start offering 100+ MB downloads? Probably not. As noted earlier, broadband users are far less patient than dial-up users. However, this does mean that we are free to explore our medium in a more relaxed state. Sure, we all have to worry about our own bandwidth costs; that is always a consideration. But it appears as though the casual audience — and I would presume the core audience to a much greater degree — is comfortable with larger downloads than we’ve ever experimented with before.
Unfortunately, these statistics do not yet measure many important things. The integration between order taking and download monitoring is minimal at the moment, though it does appear that broadband users are in the vast majority of those who eventually register the game. In addition, different channels — or referers — have different failure rates. I tried to organize the data in such a way that it most accurately reflected the casual games audience, and I believe I acheived that to a large degree. Finally, one thing I don’t have is a vast array of different titles, so these statistics are unique to users who were intrigued by Jeweltopia. However, I don’t suspect they would change much in the same family (casual).
If you are intrigued by the volume of valuable data I gleaned from my users’ download habits, you may be interested to learn that Solari Studios is planning to release this technology as a product named “trackdown” in the near future. With the capacity to monitor different products, channels, affiliates, success rates, conversion rates, and much more, trackdown will prove to be an invaluable tool for anyone who takes the downloadable software industry seriously. Please stay tuned, if you are interested in this exciting technology!
Frankly, most developers seem to spend zero time coming up with the theme of their game. This is a huge mistake: It’s your first line of marketing, no matter your distribution. (Casual portal or own site.) Just as we all wrote up outlines of our papers in school before we began putting pencil to paper, it’s important to write an outline of the specific features and characteristics of your target audience.

They’re at the top of many people’s holiday gift list, but violent video games can change a person’s brain function and desensitize them to real-life violence, a new U.S. study suggests.
I think people are far too quick to write off a web designer as “good” just because he can whip up some nice graphics. A website is a lot more than pixels, and that seems to be the first thing forgotten by people looking for a new site. (And forgotten by many that actually design the sites!)
You may have noticed the frequency of my posts on this blog dwindling over the last couple weeks. And, if you are a particularly curious person, you may be wondering why. Well, I’m hammering out the final bits of a little project which I hope to release soon. I’ve been sweating 12-15 hour days with nary a break — and I could really use one — but I want to finish it before taking any time off. Unfortunately, my blog doesn’t have the same priority as, say, going to the bathroom. (Though, neither does eating, apparently.)
Independent studio, Q Entertainment — the folks who brought us Lumines and Meteos — have announced they are going to create an MMOG. Details are unknown right now, but considering the massive investment and upkeep traditional MMOG’s require, I suspect Q Entertainment may explore alternative MMOG’s. Q Entertainment has birthed many puzzle games in the past, so we may find ourselves enjoying a puzzle-based MMOG in the future, not unlike