a black and white sign

This article is part of our Mobile Teardown Tuesday blog series. Each Tuesday, we choose a mobile app and suggest a few experiments the team might consider running to optimize the user experience. Learn more about mobile a/b testing from Optimizely.


Teardown Tuesday: Sosh

Have you ever had a friend who always seems to know what’s going on Friday night? You know, the hyper-social butterfly type who thrives on knowledge about bands and DJs in town, new places to eat or drink.

Your friend’s not-so-secret stream of socialite prowess might simply be Sosh. Sosh is a members-only app that helps people discover things to do in their local area.

Retention: Creating a loyal community

Sosh recently announced that it’s building a new marketplace to allow merchants to promote events and sell tickets on their platform. This is a big bet for the Sosh team – if executed well, Sosh could become the frontrunner in an entirely different category of apps, one that attracts both individuals and merchants.

But marketplaces are notoriously hard to create because they require the company to balance fostering a large and engaged audience while also attracting the right merchants to participate. It’s an extremely difficult balance to strike, but when done well the result is incredibly rewarding.

A marketplace needs a loyal set of users who return again and again. In other words, great marketplaces have great user retention, and Sosh is no different. They’ve cultivated a loyal following of young, urban professionals looking for interesting activities to take their minds off the 9 to 5. So much so that roughly one in six adults in San Francisco between the ages of 20 and 40 are already using Sosh.

But there’s always room for improvement, and now that Sosh’s aspirations have grown to the level of becoming a marketplace, user retention will become an increasingly important metric for them to maintain. So, for this week’s Teardown Tuesday, I’d like to walk through some simple A/B test ideas that Sosh could use to improve its mobile user retention.

Experiment #1: Sign-in Later

Sosh’s current signup flow requires you to sign up for an account immediately as part of setup. But do users really need to sign up in order to start deriving value from it? In past Teardown Tuesdays, we’ve recommended that apps which require logins prior to use consider delaying login to later on in the experience. Why? Research shows that apps lose 56% of their users by requiring registration prior to use. Users don’t want to give away info about themselves if they aren’t sure what they’re getting in return.

For Sosh, a great time to ask for a signup might be when users “bookmark” an activity they want to do in the future. When a user bookmarks something, it means they’ve discovered  something so interesting that they want to remember it for later. For me as a user, that’s real value, and it’s probably worth providing my email address for.

Here’s an example of what that flow may look like:



By implementing this test, Sosh will probably see a lot more new users checking out their product, but the real question is whether they stick around. If it doesn’t improve retention, it might not be worth it – that’s why we test!

Experiment #2: Getting Social

Sosh’s name comes from the first syllable of the word “social,” but you might not realize that after a while playing with the app. When you join Sosh for the first time, you’re automatically connected to all of your Facebook friends who already had accounts at that moment. After that, the only option for you to continue following friends is to invite them directly via email or one at a time by “following” the new friends as they show up in your “Followers” list.

A good growth hacker would tell you that viral growth requires easy, frictionless entry points into the process of friending existing members of the service and inviting guests to join. Having more friends also adds a powerful layer of social data on top of Sosh’s recommendations, promoting an even richer user experience.

Sosh could accomplish this by adding an “Follow Your Friends” banner to the home screen. Here’s an example of what that experiment might look like:


In this test, Sosh could measure retention of users who interacted with the “Follow Your Friends” banner and those who did not.

Experiment #3: Bookmark View

One of the most impressive things about Sosh is its vast library of activities with enticing descriptions and images. It’s a discovery-first experience that draws you in and can have you scrolling through “date night” ideas for days if you’re not careful.

It’s easy to collect tens or even hundreds of activities in your bookmarks list without realizing it, but bookmarks are most useful when they’re easy to refer back to and act on. Sosh does an okay job at making your bookmarks available, but the primary experience when you open the app is discovery. It says “here are some new things you haven’t seen before.” Instead, why not try, “Here are some things you’ve been wanting to do.”

An easy test to this end might be highlighting the number of bookmarked and completed activities on your home screen, so you can “keep score” and feel that sense of accomplishment.


Bonus: Phased Redesign Rollout

As mentioned above, Sosh recently rolled out a major app redesign. Big changes like this are perfect candidates for gradual or phased rollouts. In other words, rather than rolling out a single design update to everyone all at once, push the changes to a small but increasing percentage of users over a period of time and take a look at key metrics. If key metrics like retention drop, then it may be time to consider running experiments on specific aspects of those changes. The last thing you want is to make it harder for people to use your app.


Sosh is an excellent app that’s been able to grow its user base by delivering a lot of value, but user growth is only half the battle. As it sets its sights on becoming a marketplace for local events and merchants, improving retention will become increasingly important. When it comes to moving key metrics like retention, there is no silver bullet. Apps that succeed in this area are the ones that listen to their users through ongoing iteration and experimentation.

If you work on a team that develops an app, think about some simple ways you can improve your own app’s retention metrics. You may be surprised by how a few simple changes can keep your users coming back for more.