GDC 2011: Perfecting The Free-To-Play Battlefield Heroes

Ben Cousins, general manager of EA's free-to-play Easy studio in Stockholm outlined the team's experiences launching, tuning, and succeeding with the company's first non-casual free-to-play game, Battlefield Heroes, at a GDC 2011 presentation today. The experiences gained from Heroes have driven the team to launch Battlefield: Play 4 Free, which is going live in four weeks. Easy operates as its own publisher, with community, marketing, and more functions handled in-house, allowing it to operate almost independently. Battlefield Heroes started as a project at EA's DICE studio in 2007, and in late 2008 spun out as a new business unit. In December 2009, however, there was an uproar about the change in the function of the game's economy -- a huge forum campaign, a deluge of email, and plenty of negative news stories. Says Cousins, "It looked like, according to our forum posters, business blogs, email, etc... The game was obviously going to die. How come Easy still exists, and how come they are launching Battlefield: Play 4 Free?" BFH launched without the ability to buy more advantageous weapons, despite the team consulting with Neowiz, a Korean F2P company EA part-owns. Cousins attributed this to a number of factors, including the team being made of western game designers with certain expectations about what players would like; negative press garnered from a mistaken text string in Battlefield Bad Company which implied weapon sales; and assumptions about cultural differences. Promises Cousins made about a supposed lack of in-game weapon sales "are still on YouTube and come back to haunt me until this day," he says. 
The game launched in June 2009, and July 2009 was biggest month to date for user volume. However, after launch, the team faced a "really big problem ... it was clear we weren't making anywhere near as much money as we forecasted." Average revenue per user (ARPU) was targeted at 50 cents, but the real return was 25 cents in the first month. "We'd used Korean benchmarks and were even a little conservative." The team identified four key performance indicators -- monthly ARPU, monthly active users (MAU), monthly conversion rate, and average revenue per paying user (ARPPU). 
In the July 2009 KPIs, ARPPU was $20.25, but the conversion rate was a mere 1.29 percent. "Conversion rate was the issue -- clearly failing one on of the KPIs allowed us to really focus our work." The team discovered a class of players it termed "'peacocks', who wanted to show off their clothes." They ran a survey to find out what players wanted to buy, and contrary to expectations, advantages in battle were high on the list. But the team still didn't feel comfortable implementing those, and they were added to the bottom of the backlog. Change was slow; Cousins asked for more staff, but was told they would not be forthcoming until they made more money."The problem was that my bosses at EA didn't know if free-to-play would work in the Western world ... So they were reticent." In November, Cousins said they were making progress, but still avoiding advantage items. But in early December, EA cut 1,500 developers across the company, which made the Easy team scared they would be cut too. EA decided that the game was "too free," Cousins said; metrics showed that free currency earned through play allowed players to maintain multiple valid characters for totally free. The decision was made to tune the game so that you could maintain one character with one weapon if you played a couple hours a day. The team dropped the prices of permanent weapons because "we wanted to increase the accessibility of those while decreasing the accessibility of free weapons." The team also finally introduced more powerful weapons for cash: three with bigger magazines, five with a slightly better chance of a critical hit, two which did slightly more damage, and one with a recoil reduction. Cousins estimated "a 10 percent increase in power for each of these cash weapons." The first new weapon launch -- shotguns -- resulted in a 64-page forum thread the day they were introduced. The mods created a new thread which swelled to 50 pages and 1000 posts in a day, and later to 250 pages and 4000 posts. The team quickly "got nervous -- but lucky we were running an online direct to consumer business and we had the data," Sessions said BFH had good MAU and ARPPU, but bad ARPU and conversion prior to the changes, but there was a 100 to 200 percent jump in daily revenue overnight with the new guns. DAU didn't change whatsoever, and while ARPPU dropped during this period, there was a "a really dramatic increase" in conversion -- three times as much. "We were still under a five percent conversion rate daily," says Cousins, but still much happier. 
 In other words, "there was a mismatch between what was being said on the forums and what was happening in the data." So they analyzed the forum. The stats: 78 percent of their users never touched the forums. 20 percent read at least once but didn't post. Two percent read and posted on forums, in proportion of the game's total user base. A look at the spending habits of forum posters to determine if they spent or not revealed that their average spend on the game was $22 -- or more than 10 times the average user. In other words, "there seemed to be a disconnect between what they were saying -- 'I will leave and never spend a penny' -- and what they were doing -- sticking around and spending a lot of money," Cousins says. Given this data, the team decided to "use forums as an alert and then do a data-driven examination" of potential problems. Did this have a negative effect on the game? Says Cousins, "There is no long term effect from pricing on registration." Moreover, "there is no real long term impact in trend of churn." Meanwhile, "gross funding revenue has gone up a lot." "Battlefield Heroes is very profitable now and our projections for next year have us running at a 50 percent margin," cousins says. And with upcoming launch BFP4F, " we aim to make this the western world's biggest free-to-play client game." Says Cousins, "Common assumptions about the types of virtual goods it's OK to sell in games are incorrect. It took our fear of losing our jobs to take a leap in the dark." He analogized this to amateur sports -- you can buy better equipment in cycling or golf, but everyone accepts this and still plays and competes. "Sometimes the guy who killed you with a good weapon was a good player and he would kill you anyway; sometimes he's not a good player and the free players enjoy ganging up on him and killing him anyway," says Cousins. Since the game's combat loop is 15 seconds, free players can devise effective strategies, Cousins says. "It's actually what drives all economies whether you like it or not." In conclusion, "Stop trying to define user behavior... Instead look at what consumers want it to be."