Albert Reed is the CEO and cofounder of Demiurge Studios. Follow the entire series here.

Every event in Marvel Puzzle Quest includes a custom-designed card pack. An enormous amount of discussion and effort goes into making these designs. We haven’t perfected it yet, but there are some things we’ve been doing consistently that seem to improve performance.

Marvel Part 4Card-pack design

  • Help event performance: Card packs are carefully structured to give players a clear path to improving their performance in events. Frequently the desired design of a card pack ends up impacting the design of the event itself. We will add or remove characters from an event to cater to the needs of good card pack structure. Focus on the positive: In our language describing the packs, we always make sure to phrase things in the positive. Rather than saying “At least two Powered-Up characters,” we would say “Two or more Powered-Up characters”.
  • Multipacks eliminate risk: We increase the value of the multipack by promising the player that they’ll get one or more of the most-valuable characters in that pack.
  • Fun designs: When designing an event and its corresponding card pack, we always look for a fun theme. Packaging our cards in a logical grouping enables us to articulate the contents of the card pack with less text.

Sales

Offering more in-game currency for less is a much-discussed topic in free-to-play game design. Some games never do it, and others do it constantly. In our survey of games, there seems to be a close correlation between successful games and those who use it sparingly or not at all.

  • Breadcrumb: We lead the payer into the sale with a tasteful tweak to the store button that reads “Bonus.”
  • Players will buy: They work. Sales, regardless of size, seem to at least double our daily take.
  • Expect a hangover: Like any overly fun night on the town, the next day is painful. Expect to see a meaningful dip in revenue in the day(s) following the sale.
  • Clearly communicate duration: Within the currency store, we make clear how much time is left in the sale.
  • Measure success by conversion: One way to decide if a sale was successful is by the number of players who converted for the first time during the sale. The theory being that if players have already monetized, the sale probably just borrowed from future revenues.
  • Design sale around instrumented excess currency: Keep an eye on how much currency is flowing around in your economy. If you detect that there is a glut, chances are the sale will simply convert players who were likely already inclined to buy.

These two optimizations closed out our Q4 2013 improvements. At this point in the game’s development, we’d seen many days with ARPDAU that was double where we started and spikes during currency sales of considerably more. Perhaps most-encouraging is the curve of spending for players showed no sign of slowing down: the cohort of users who started playing in early December were spending as much on day 30 as they were on day 7. In the next part, we’ll discuss the specifics of how our “Subscription” model performs and a couple of new features we’ve added into the game to enhance that performance.

Demiurge CEO Albert Reed.

Above: Demiurge CEO Albert Reed.

Image Credit: Demiurge

As cofounder and CEO, Albert Reed has held the reins for over a decade at Demiurge Studios, a game development house focused on innovative, high-quality games for all platforms. Find him on Twitter at @almnop.