In his GDC Online talk Tuesday, Ethan Levy of BioWare San Francisco revealed discussed Facebook game Dragon Age Legends' promotional campaign got a lot of likes, but not much else.

Almost 40 percent of the likes the game got on Facebook were generated
by promotional efforts -- counting likes appearing within two days of
each major promotional effort.

Three press releases, distributed by EA's PR team, got traction from
blogs like Destructoid, Kotaku, and IGN -- which was due to those
efforts and also the appeal of the Dragon Age brand, he thinks.

One tool that was invaluable was the creation of key art for the game.

"We had three posters made to help communicate the three main ideas of
the game -- assemble your party, build your kingdom, and be legendary.
What was really important is that these posters" could be used in
promotions, such as IGN's front page, said Levy. They also helped
communicate the message directly.

The same artist who produced the artwork also produced a two minute
trailer -- which got over 100,000 views, Levy estimates. "That trailer
was served to a lot of websites," he said.

As far as their creative tactic for it, "It was a kind of taking a page
out of the BioWare book," Levy said. " They don't show gameplay; they
show the kind of mood the game is trying to create, and give a little
backstory."

Levy ran a promotion where he implied something would be given to the
community at a 100,000 like target. He had to skirt the issue because
Facebook policy does not allow direct incentivization of likes.

"People really liked it, the art was really nice," but in the end "I
don't think it had anything to do with" getting to 100,000 likes, he
said. He thinks it was a natural process based on the overall promotion
of the game.

"If you are making a game, there's a lot of art around you all the time,
and all you have to do is put it up on Facebook and people will like
it," Levy suggested. "If I think it's awesome, chance are fans will
too."

"We established our voice on the wall, and I'd say that voice was pretty
playful," he said. He mentioned ideas such as pretending the beta keys
were guarded by monsters -- writing promotions in the style of the game
fiction went down well with fans. He also responded to fan comments with
content on following days, creating a back-and-forth.


"Competitions had the highest engagement," said Levy. "We let our fans
give us the war cries" for some of the characters in-game by commenting,
which resulted in 300 unique entries, six of which were used in the
game.

"People who love the game love naming things in the game," he said.

On the other hand, "Blog posts got a medium response ... and were very
time-consuming to write... Those blog posts did not get the same sort of
response that a piece of concept art or a quick piece of news did."

Videos sometimes worked, sometimes didn't; they're also time-consuming.

"The things a producer does, like write on a whiteboard and prioritize
features, can be turned into a content your audience cares about," on
the other hand. Showing a bit behind the scenes to tease new features
works well, Levy said.

He also tried a beta key giveaway in San Francisco -- 50 people showed up in two hours.

On the other hand, cross-promotion for other EA products on the Dragon Age Legends feed "drove a lot of unsubscribes. Your fans do not like it."

The bad news: this built a huge early audience for the game, "but they
didn't stick. We launched the game too early. The game was not in a
state to support that momentum. We shot ourselves in the foot by
sticking to that marketing date."

"We could have used a much longer closed beta where we fixed bugs,
listened to feedback, built the community, and made more content," said
Levy.

"In free-to-play, a few Kotaku articles and IGN front pages do not make a
hit game... All you've convinced people to do is show up and install
the game and maybe play the game for one day."

"How I picture a free-to-play launch now is to launch in phases and have
a clear target of when you move from one phase to the other," he said,
referring to the idea that he ought to have multiple phases of closed
beta to make sure the game would play right at launch, and store up
content -- not build almost all of it post-launch.