DIY.org: Case Study

A mission-project with a bittersweet outcome.

DIY.org was started as a project to create a place where kids could teach each other skills, show off their achievements with badges, and gain relationships and attention in a safe space. DIY was originally founded by Vimeo co-founder Zach Klein, Isaiah Saxon, Andrew Sliwinski, and Daren Rabinovitch in May 2012, but apart from Isaiah (who continued to produce illustrations) the other founders exited to work on other projects.

At that time, most of the skills content was sourced from free content available on the web or contributed free by members. When I began working with DIY.org, the small team had about 12 months of runway and limited revenue, most of it from selling physical goods associated with the otherwise free site. There were significant compliance needs moderating content contributed by kids, and no indication that marketing the existing product more aggressively would produce a sustainable business.

The CEO Zach Klein was and is a visionary who believed in the idea, and I worked with him and Chalon Bridges (Director of Education and Content) to develop a set of paid courses for kids that could be a more valuable version of DIY, and represent a path to growth. I took on the role of product manager and product designer, with a team of four.

v1: Camps

The first version of the concept was called camps.diy.org, and was launched in three weeks with just myself, two engineers, three moderators, and six “counselors” responsible for the courses.

We launched at a price of $20/month as a subscription (with a 1-week trial period). Parents signed up and kept track of the camps via a web app. Kids used a new version of the DIY iOS app to interact with the courses.

Each of the 6 launch courses content was to be new video lessons posted every day, and counselors would have live interactions to encourage engaged kids.

We started just as the summer was also about to start, to position ourselves as a helpful tool to busy parents who might need help keeping their kids productively engaged in gap times or longer periods out of school.

Initially, there was a strongly positive response; we signed up almost 1,800 kids in the first week, and 1,600 of them converted through their first month. In subsequent months we saw 24–30% growth and churn dropped to less than 5%.

By the end of the summer we had almost 10,000 subscribers. We knew there would be a large dip after that as kids returned to school, but the falloff was a lot less than we expected. We believed that we had a model that could work with some adjustments.

v2: DIY Courses

The second version of the product integrated the courses into the free site and created an upgrade path for free subscribers. After that basic block-and-tackle work, we set out to create a sustainable business.

In the design, I focused on making a friendly, ultra-simple, ultra-clean set of affordances that wouldn't “talk down” to the kids and had all the energy of the playful adults that loved the product they were making.

I preferred to work in pencil and paper, revising the concepts continuously in response to feedback from team members and kids (including my daughter, who was a fixture in the Minecraft course).

We decided to charge a yearly, flat price of $99, and transition the courses away from live interactions with teachers to a new moderation model supervised by our moderation staff.

That would reduce the costs of producing the courses (and we could invest in the quality of the videos instead), and the courses could scale up to the numbers needed to make the business model work.

Roughly, we figured that we needed to grow to 3,000 kid subscriptions, 100 kid-moderators, a cost per acquired customer at around $10, a month-over-month subscriber growth rate of about 20%, and indications of an average customer lifetime of around two years. That scalable model would (after 1 year) generate about $300,000 of annual recurring revenue in the first year. As long as we were able to do that without growing the size of the team significantly, that would achieve a significant valuation for the company.

We met with venture funding firms and outlined that plan and had a good response and indication that performance like that would generate additional funding or acquisition interest.

The Moderation Model

Our ability to scale depended on our moderation model working, however! For philosophical and economic reasons, we really did want DIY.org to be a place where kids were the major moderators and helping teach other kids. We had seen it work in the free product, in Camps, and we believed it would work for Courses.

We worked intensively and iteratively to launch and prove-out our moderation model, where kids with longstanding accounts and skills would be encouraged to become moderators for the relaunched courses.

To track the kid's progress within the full set of courses we introduced “XP” that was earned with each course completed, and provided ways for kids to have the content kids made in the courses featured and win currency in the form of “gems”, which could be used to buy merchandise.

In this scheme, kids taking a course would become candidates to be a moderator for a new section of that course automatically. They would earn our currency and XP, and be able to show off with special flair on the platform their status.

As kids completed a course, it would feed into a set of other courses naturally, as part of an overall “tree” of content. We also invested in video production for the courses, making them much more entertaining and rich.

Promotion and Growth

In the initial release of this new version, we had a burst of interest and some indications of ongoing growth, but also higher churn than we thought was acceptable. At the same time, we signed a promotion deal with Cartoon Network to run a contest that produced a large influx of new subscribers.

Acquisition and Epilogue

As we were making changes to address the issues and growing quickly, the board made the decision to accept an offer to purchase the assets of DIY.org by LittleBits.

While this was a version of success and had a good outcome monetarily for the team, all of us were somewhat bereft that we would have to put our plans on hold until the acquisition was complete and it became clear what the new parent company would decide for the product. Ultimately, the team was not part of the acquisition.

All told, I spent almost exactly a year leading the product and still have a strong amount of affection for the entire team involved! We all still love the product and put everything we had into making it succeed. I learned so much there, and I am grateful for the experience!