Cluster’s Evolution

Refining our focus as users stretched the product


Brenden Mulligan is a co-founder and designer at Cluster, a web and mobile app which enables users to create private social networks around interests and experiences. Brenden previously created and sold ArtistData and Onesheet.


When we started building Cluster, it was a seemingly simple app that allowed friends who’d shared an experience together to exchange photos. It was born out of a simple frustration that the tools currently out there to accomplish this task were clunky and flawed. We knew we could do better.

So Taylor and I launched the initial (hideous) MVP of Cluster to solve that problem.

https://twitter.com/allthingsd/status/306829273807859712

At first, it was just a side project and we were excited that people thought it was cool. But almost immediately, we started to see interesting behaviors evolve that we didn’t really anticipate.

https://twitter.com/tconrad/status/304768357910065153
https://twitter.com/hunterwalk/status/326721970320314368
https://twitter.com/emilyolson/status/332476500370677763
https://twitter.com/mistersterling/status/342081036911509504

It was clear we’d built something more interesting than intended, so we turned the side project into a startup and have spent the past year working on it.

5 Use Cases Evolved

For the rest of 2013, we continued iterating and watching how people were using our app. We considered ourselves in the public-beta phase of the product, where we weren’t as concerned with growth as we were with usage patterns.

During this time, we started seeing distinct use cases evolving.

Everyone Shares, Everyone Cares

We designed Cluster to help a group of people who all had a similar interest in seeing each other’s photos. They used the app to collectively to build a collaborative photo album after an event or vacation . This has always been a heavy use case.

Pros: The app was designed and optimized for this. Everyone’s shared interest in the photos makes most of the features we’ve created work like they should.

Cons: The albums tend to expire shortly after the event has passed, diminishing the value of the app and making the user less likely to come use it again. If the user goes to a lot of events, this is less of a problem, but not that many do.

Photo Backup

People are taking more photos than ever, and many times they want to get them off their phones. This has led to many users uploading photos to Cluster that they never intend to share with anyone (about 1/3 of the total photos uploaded). Users create a series of albums to organize and backup their photos, yet they never invite anyone to these albums.

Pros: These users find great value in Cluster and would be upset if the app disappeared, definitely a good sign for the product.
Cons: These users provide the least value because they don’t invite others to the service, yet account for a large cost because they tend to upload an enormous number of photos. Plus, we didn’t build a photo backup service. There a many that do a phenomenal job at this, and some that tried and failed.

Gimme Your Photos

Our app works very well for brides and grooms who want to collect all the photos from their wedding. This is slightly different than the first use case, because the whole invited group — who might not know each other—doesn’t have a shared interest in the photos, just the bride and groom. The purpose of the album is more of a one-way dropbox, so the album creators can get the photos together in one place.

Pros: Creators of these albums tend to invite many people, which leads to good viral growth of the app. They also tend to have a ton of photos uploaded, which gives a really wonderful experience for the creator.

Cons: The people who are invited don’t necessarily care about seeing all the photos (or push notifications) from the event. Also, when they’re invited to an album like this, their perception of Cluster locks in as “an event / wedding app” and it’s hard to ever get them to create an album on their own. And again, after the event is over, the album expires.

Limited Broadcasting

Sometimes a person or small set of people take a lot of photos they want a slightly larger group — but not everyone they know — to see. An example here could be a couple that goes on a vacation and wants to share all their photos with their family back home. While they might post 5-10 on Facebook or Instagram, they post hundreds into the album on Cluster.

Pros: Tons of activity as photos are added. People who are shared with usually are very very excited to get app alerts and provide lots of likes and comments for their sharing friends. Albums tend to be very social and active.

Cons: Since most of these albums tend to be tied to an event with a start and end, these albums usually expire relatively quickly and the invited users become disinterested in the app when the activity wanes.

Private Group Sharing

The final and most surprising use case for Cluster has been around private sharing. People generally share different interests with different groups of people, and it’s hard to separate how you share with those groups on existing social services. Within Cluster, users are building mini-social networks, where a group comes together around some kind of interest or connection, and shares on a daily basis with no pre-determined end date. As our users describe it, “Google+ Circles done right.”

We’ve seen some powerful examples of this around new parents, who create an album and invite just the family members who want to see a million photos of the baby, instead of over-sharing to Facebook. We’ve also seen this heavily used with college students (especially fraternity and sorority members) who take a ton of photos that might not be appropriate for more public social networks.

Pros: These users return day after day to share and communicate with each other in a private setting. The interactions are somewhat similar to group chat apps, except the format is more similar to a social network. Photos are persistent, not ephemeral like SnapChat, but the space they’re being shared in is private and only accessible by trusted, invited friends. This lets people be themselves. We see huge, continued engagement here and the people using it consistently derive a lot of value from the groups.

Cons: There are a billion social networks. And for this use case to really work, the initial user has to correctly set up the group in a way where invited friends or family totally understand what the space is for.

Focusing

It’s great when users find lots of ways to use your app. Sometimes they unlock things you never intended. The important part of the process is to listen and learn. That’s what we’ve been doing for the past year at Cluster. Through this process, we’ve added a lot of different features to the product, which have given users the ability to use our app in really interesting ways. However, now it’s time to focus and make sure that the most powerful use case is easy to discover.

We’ve spent the past few months studying users and redesigning the app to focus on private social sharing. We think that that’s where the puck is going, and we’re experimenting a lot with how to do it right. We face major hurdles, including growing a user base with an inherently private product with limited viral loops.

It’s not going to be easy, but the compelling use cases we’ve seen are too hard to ignore. We’re excited for the challenge.


Thanks for reading this! If you got value out of this article, I would really appreciate you hitting the recommend button below. Connect with me on Twitter @mulligan with any comments / thoughts.