Lean Startup example: kidyos - in search of a product/market fit for kidyos

First, a bit of context about the project.

 kidyos (http://www.kidyos.com) is a web-site for parents and thier toddlers. The idea started a few month ago, with the main hypothesis being:

    (a) there's  a lot of  high-quality kid-friendly online content  that is suitable for this age group (think youtube clips, mini-games, etc.)

    (b) there is value in aggregating this content into one destination site so as to make it easy for parents to access & discover new suitable content  

    (c) parents will pay for this value, to a point where it can become a viable business.

Discovery Phase 1

the genesis of the need for kidyos is very much personal experience -- the little media we expose our daughter to is online, primarily through YouTube and a few other sites (we do not have a TV at home). There are tons of stuff for kids out there but I always struggled maintaining a current list of the things she likes (preferences change quickly at this age) and I'm always concerned when exposing her to new material (you never know if it's safe for kids until you watch it yourself). So I definitely felt the need, and assumed others may too.

I had a few reasons to work on this project beyond "building a business" (though that was and still is the goal) which occasionally take me off the pure lean-startup path (such as building something  for my daughter ... or re-learning how to program after 10y away from the craft), but ultimately my goal is to explore the kidyos concept using lean methods.

Luckily, I had no shortage of potential customer leads. Many of my friends, and friends-of-friends have children in this age-group, so finding relevant people to talk to is easy (one of the many advantages of building something you are also a user of). My first set of customer discovery discussions where to simply ask parents what sort of media they have their child watch and talk about the experience and difficulties they might have, gradually suggesting an idea like kidyos if the need exists for them. I did not go through this scientifically unfortunately, but after speaking with about a dozen parents or so, I could see that a large percentage (about 70%, though I now know this number is skewed upwards because of the demographics of those surveyed) do in fact rely on the web, to some extend, for their child's media entertainment. The conversations were illuminating as there was a lot of communality between parents: use adhoc methods to maintain their favorites, concerned about safety, struggle to find new material when needed, etc. In short, good indications that the need existed and good feedback on the solution as described.

In parallel to these conversations I was developing an MVP to test my hypothesis (and entertain my daughter). A simple flash-based player that uses  an even simpler database with a bunch of YouTube clips I collected over the years. The player basically allows you to navigate through the database, find the clips you want and play it (in retrospect I spent too much time on this MVP, but that's a different story).

Discovery Phase 2

I then made an offer to each of the parents that expressed the pain to participate in my 'early release' and see if it's useful. This was useful as it filtered out those that didn't feel the pain enough to bother, or were just being polite or for whatever other reason were not a real 'earlyvangalist' candidate. At the end of this process I had 5 families using kidyos on a regular basis.

At this stage, I focused on understanding usage patterns better. With such a small user-base this was easy, I literally spend a couple of hours every week sifting through access-logs to determine what people where doing. I spent about 6week with this pathetically small user-base, using a really bare-bones, horrible product. I was determined to stay in this mode until I see marked improvement in two metrics:

    (a) increase engagement to a point where kidyos becomes the default gateway for their child's online entertainment ('measured' this by looking at the logs + routinely asking "how many times did you use the computer with your child this week NOT on kidyos")

    (b) wait for parents to reach out to me and ask if they can share kidyos with friends they know. I deliberately did not request or mention this myself, as I saw parents proactively requesting to do this is the most practical proxy I have at this time to gauge their perceived value of the product (people recommend only what they really like).

During this period I routinely changes the product in various small way, often surprised by what was effective for achieving these goals. For example, I ended up spending significant efforts in making the UI simpler and more streamlined. I incorrectly assumed that, this being an MVP, my users will see past the ugly interface and workaround any usability issues.. and that the way to improve the product-fit would be by identifying more capabilities/content-types they require. This was not the case. Turns out my early-users are not particularly tech-savvy so they would often get stuck, and the context in which they use kidyos (usually with a 3-year old jumping anxiously around them) requires the interface to be straightforward, simple and fun to navigate. Several weeks later, this is still work in progress - the site is still not there in this respect.

It is worth mentioning, that in this entire process I "lost" some of my users, at which point I'd make the effort to recruit someone else instead.

Discovery Phase 3

After about 6weeks of work I got to a point where I was satisfied with both metrics: parents in my user-group used it regularly, and I got 3/5 of them to ask me if they can recommend to friends. I also spent this time making improvements to the underlying infrastructure, dabbling with AWS and so-forth (re-learning software engineering after 10 years results in a lot of mistakes that need to be fixed :-). I built a very simple registration function on the site (so people can signup themselves) and "launched".

 I asked my existing users to recommend kidyos to friends, did some more personal outreach myself and started sending out emails to several parenting lists I've collected. I've been doing this very slowly, so I can test the effectiveness of my new-user funnel (from intro emails thru the registration page and onwards) and reviewed and reach-out to customers for interviews and questions.

The feedback has been very illuminating, for better and worse. I know much more about the type of families that see value in kidyos, and how I can make it more useful for them. I have a list of things to change and improve, and a bunch of ideas to test (while the number of users is still very small, it's at a point where it's practical to test ideas and observe users' reactions). I plan to continue and do that while I experiment with more sources of incoming traffic (adwords, etc.). My goal is two-fold: Increase the user-base size to make it easier to reach conclusions that are statistically more viable and improve the utility of the product and the resulting user engagement levels. 

Assuming I reach that goal, the next step after is to experiment with pricing. Throughout these discovery conversations I've been testing the business model I have in mind for kidyos and have gotten positive reactions. But I remain skeptical until I actually see it in action. If that turns out successful, I can quite confidently evaluate the business merit of the project and determine how to take it forward.

Conclusions / Observations

    * A interview / casual conversation with intended customers is a great way to start. Really helps sharpen thinking and identify promising leads for initial users.

    * That said, one can easily get stuck at this stage indefinitely (I was lucky to have a practically unlimited pool of highly qualified leads to lean on, and a reasonably sized list of leads can take a lot of time to build). it's important to get users to start using something as soon as you think you've identified a seed of a community with common needs. 

    * There is nothing more encouraging, enlightening and motivating than seeing actual users using your product - that reason alone justifies releasing something as early as possible. 

    * While I often feel I could move faster, I find it very valuable to build a small community of users and stick with that group to validate  ideas and hypothesis - both the essence of the business and any sub-features/capabilities. Been doing this for a couple of months now and still consider myself in this discovery phase. 

    * Going through the process takes time. There really aren't any good shortcuts and it's especially challenging when done on a shoestring. It's often tempting to do it the "old fashion way": convince a bunch of people (i.e. engineers and investors) that the idea is the next big thing, assemble the team and the resources and develop the current vision :-)