Browser Demo = WIN
Kongregate also brought in a large amount of revenue, not only as a large portion of our direct sales, but also through Kreds transactions. Our browser-based demo was our ticket out of obscurity, and flash game portals drove lots of sales. Even after flash portal traffic died down, the browser demo continued to be our secret weapon.
This is because of the "EXE barrier." To install a downloadable demo, a player has to click a link, download a file, navigate to whatever stupid folder their browser saved it to, double click it, be warned that the file will destroy their computer, install it, wait for it to install, and then run the game.
As I've said in my article Piracy and the Four Currencies, even if something is free in terms of money, it can still cost people time and pain-in-the-butt. A browser-based demo is one click and you're playing. No money, minimal time, minimal pain-in-the-butt.
This has obvious benefits in terms of demo-to-sales conversions, but it doesn't stop there. Journalists, bloggers, Steam/GOG reviewers, and other gatekeepers will play that demo, too, and the EXE barrier affects them all.
So, if there's one thing you take away from this article, it's this: make a browser demo. Then, cap off the player's positive demo experience by letting them export their save progress. Remember -- time is more important to some people than money. Some players might be happy to give you $15, but not if it means losing the two hours they sunk into your demo.

Browser games have traditionally been Flash-based, but today there are more options than ever before. Unity, Java, HTML5, and Haxe are all great alternatives, and Kongregate supports all of them. Plus, Unity and Haxe can export directly to Flash, so you can have your Flash-based cake and natively-compile it too.
Remember that little game called Minecraft? A chief key to its viral success is that its demo can be played in a browser.
Seriously, I'm going to keep saying this until I'm blue in the face.
Apples & Oranges
The pie chart alone isn't a fair comparison, however, because we've been selling direct for much longer than we've been on Steam, et al. Therefore, we need a chart of sales over time.

Direct sales peaked early, then trailed off. The only spikes after the initial launch came from Rock, Paper, Shotgun articles, being featured on Kongregate, and participating in the Because We May sale. We saw another small spike when Gold Edition came out, after which sales ticked up marginally, but that was it.
Without new sales channels, Defender's Quest's run would have finished.
Daily Deal Delivers the Dollars
The Steam Launch was huge. We launched at 33 percent off the final price of $14.99 for the first week, and later participated in the Autumn sale, but the best was yet to come. When Defender's Quest was selected for the Steam Daily Deal, we made nearly half of the game's entire pre-Steam revenue in a single day.
The daily deal spike is actually slightly under-represented, because the 24-hour sales period was spread out over two data points -- the total take being nearly $34,800.
GOG gave us a similar one-day promotion, a "GOG Gem" feature, on January 23, for 60 percent off, which was another nice, single-day treat.
GOG is Great
GOG was overshadowed by Steam, but still a major source of revenue and matched our take from Kongregate Kreds in a fraction of the time. Shortly after launch, GOG embedded the browser demo in the game's news page, driving further sales.
Even better, Steam and GOG continue to have healthy long tails. A year from now we expect Steam and GOG to account for over 75 percent of Defender's Quest's lifetime revenues, especially if we get to participate in future sales.
Desura, Impulse Disappoint
Desura and Impulse barely register on the above charts, and their combined total was under $1000. Had players not been able to buy the game on GOG or Steam instead, I suspect sales on these platforms would have been higher. I know some developers have made good sales with these platforms, so I wouldn't write them off entirely, but our experience wasn't particularly lucrative.
|
Short version:
There's no magic bullets when it comes to marketing, we could just have face-planted or struck it rich with either model. We liked this model and it seemed to work.
Long version:
http://www.gamasutra.com/blogs/LarsDoucet/20121115/181659/Free_2_Play_and_the_Fo
ur_Currencies.php
This article is exactly what I've been looking for, actually, thanks for posting it!
1) Reach the end of the demo
2) Quit
I'm not sure if this could be optimized, it just seemed natural to me.
Off-site, the kongregate version directs people back to our site. The game can detect what site it's running on for when small portals "steal" the swf and upload to their sites.
The kong premium version has the same content, just with compressed assets so it loads a little faster online, and it has a secure link to download the full version of the game for your desktop as well (uses the same update server we use to deliver auto-updates).
The Steam version is compiled in Adobe AIR and features numerous native extensions to both integrate with Steam as well as provide desktop-friendly features like native fullscreen resolution switching, etc. For mac and windows this uses "AIR captive runtime" so you don't actually have to install AIR, it just packages the dependencies as local files, but linux does require the user to install AIR - in that latter case we wrapped it in an easy installer shell script that does all the work for the user.
Thanks for putting the time into putting it together to help everyone else out!
Thanks for the article, Lars, fantastic take-aways here. We're an HTML5 studio, so having a web-based demo is done and done for us. So hopefully we'll be able to apply some of your lessons to our next game. Cheers!
Cheers!
I'm not suprised by the games success. I do remember your 1st article about it here at gamasutra, i said it should be on steam, its better than alot of the games they sell there.
Also it's nice to see your still DMR free and working with the guys from GOG.
Way to go!