Interview: How Valve hopes to improve discoverability on Steam - MCV

Interview: How Valve hopes to improve discoverability on Steam

'Giving customers more options means giving developers more places for their product to make a connection,' says Alden Kroll
Author:
Publish date:
Alden Kroll Valve.jpg

As large numbers of new games continue to be released on Steam, Valve has revamped its storefront and discoverability in an effort to improve the shopping experience for consumers and alleviate the concerns of PC game developers.

To find out what the new changes mean, which you can read here, Develop spoke to Valve UI designer Alden Kroll to ask effect the new features will really have, and why they have been made.

How will this help smaller studios and indies improve the discoverability of their games? Is this the result of discussions with developers?
Pretty much everything we do comes as a result of our conversations with developers and customers. In terms of how these changes will help smaller studios and indies, we are attacking that from multiple directions. The first and easiest step was to open up the front door and create room for more placements. But the more useful part, we believe, is taking the extra step of having new areas populated based upon a number of interesting criteria. These criteria range from your own defined tastes, player data and game ownership, Curator reviews, plus things like the Recommendations Feed.

At the end of the day, New Releases and Top Sellers are interesting categories. But that isn’t the only way customers want to shop. And giving customers more options means giving developers more places for their product to make a connection.

What can developers do themselves to make the most of this new system?
The same things they’ve always needed to do to succeed: make great games and get the word out about your game.

What do changes to discoverability mean for your approval process for games? Is Valve putting an end to its curation of the platform, and perhaps to Steam Greenlight?
Our old problem was throughput. As soon as we started increasing the number of titles being published, we knew discoverability would be the new problem. With an increasingly large number of new, interesting games getting made, we foresee our throughput continuing to increase as well. These updates are designed to utilise community feedback to help identify the right games in the Steam catalog to highlight to each player.

You talk about utilising gamers' playing/buying behaviour to surface recommendations, but how deep do you plan to go with this? If there's a player that rarely gets far into long games, would you recommend shorter titles? If a player is nearing completion of a set title, would you use that opportunity to suggest DLC?
Like everything we build, we’ll be iterating on the way recommendations work and what factors we incorporate into those algorithms. There are many great opportunities like the two you point out. But we also know there is danger in trying to make assumptions of a player’s intent based on their behavior, so we’ll continue carefully down this path.

Does the personalisation of the home page run the risk of making players unaware of some interesting releases outside their usual tastes?
The Recommendation Feed at the bottom of the Home page is designed to prevent that exact problem. It’s looking at all your other preferences and placements and making selections for your consideration based upon “what you may have missed.” Additionally, we still feature some top-sellers in the large banner on the home page and still have lists of new releases and top-sellers.

What else are you looking at to improve discoverability on Steam?
For now, we are most interested to see how these changes perform in the wild. Like any other major release, there will be things we missed, and things we need to adjust. Once we get a handle on those, we’re certain there will be more to work on.

Related

2_Steamloo.jpg

Valve on Steam

The astoundingly popular digital platform has revolutionised the PC industry, but how? And what next?