Another Month, Another Startup

August 31st saw the launch of our second micro-startup in two months, ScreenSponge.com. Our first one month startup taught us that we can work together as a team and build software. The unanticipated success that came from the launch of our first project suggested that together we may be able to build things people want to use, and clarified the counter-intuitive axiom that people want to pay.

We had a slow start, not deciding upon a direction until a number of days into the month. We were two team members down, our UI guy taking a long overdue holiday, and one of our star techies tied up with other work. A focused, although inebriated discussion between Craig an myself circled the shared problem of tracking the TV shows and Movies we watch, have, and want. Our vision was for some kind of "TV Guide 2.0", where, rather than an authoritative guide informing the passive users when they have permission to consume media, we envisaged an active ecosystem, where the system and the user actively facilitated an array of media consumption channels, not limited to TV, Cinema, Video, and the Social Network.

For example: You want Lost Season 02, Episode 09. The system informs you: it is on Channel x tomorrow at 5pm, your friend Tom has the entire Season on DVD, your local video store has it on line for $, youtube has a clip, amazon has a special today for... so on.

Craig and I quickly came up with the tag line "connecting the humans with the media they want to consume" and annealed it back into the more digestible "connecting you with the shows you want". We spent a grueling afternoon brainstorming the name, painful because every crappy name we and my housemates proposed was taken. I'm sure the project was better off being conceptualised in such depth and in having an identity so early on.

Buzzing with this vision and the anticipation of funneling, scraping, and feeding data from all over the web into our system, we set about designing the database entities and relationships. This was by far our biggest mistake, although importantly shaped what the webapp became.

It is safe to call Craig an I engineers, we think in systems. We took what we thought was an exciting vision, with a clear and unfulfilled need in the market, and jumped straight into the back end, as far away from users and functional interaction as we could. Fair enough too, this is how we were trained to build business systems. We burned a week on nutting out a functional ERD in semi-isolation, which, after one critical skype call was scrapped for the simplest, denormized, most elegant solution that totally nailed our needs.

At the time, we patted each other on the back, happy in our lesson learned regarding the simplicity of design, although our systems focus unwittingly continued to hurt our progress and dilute our vision.

With the infrastructure already in place from our previous project, we quickly fleshed out the screens of the application. We made the design decision to allow users to take ownership of their shows, rather than push a ground truth. We focused on allowing users to name a show whatever they want, write a description meaningful to themselves, a review, a cover, events, everything. The adaptive system promoted micro-communities around shows (members with discussion), and allowed the popular variations of those shows to propagate to new users (all very Darwinian).

Two weeks in and we had a suite of beta users, adding and curating shows, and we were feeling pretty happy about our progress. Given our stringent time constraints, we focused on users managing their shows, and passively interacting with friends shows lists, a long way from our original vision.

At around this time, our UI guy (Matt) arrived back from his holiday and started using the site. Although he'd been online from the beach (totally the wrong type of surfing), he'd been out of the development loop and viewed the webapp in the context of what was initially planned. "Where's the connecting part?" "How are you connect me with shows I want, when I can't even demarcate wants?" (my words)

Thankfully, Matt drew attention to the fact that we were not only ignoring the user experience of the application (!), that we'd forgotten our vision.

The very next day, we had a totally new application. A rapid redesign, the addition of three simple intention/possession states on shows (seen, have, want), and the interaction and presentation of friends shows through the states resulted in an application amazingly Craig and I (the two core developers) found immediately useful. The second we flicked the switch on "shows you want your friends have" and actually saw that each of us had shows the other wanted (that we previously didn't know about), we knew we were onto a winning idea.

We spent the remainder of the month focusing on "show possessive states" and the social graph. We added a "request from friend" feature, show hierarchies for TV series-season-episodes, discovery features, filters, search, and made show trading and acquisition amongst friends "the one thing", the applications focus.

Interestingly, we broke one of our primary rules in our company vision, that is our application is monetized via advertisement rather than directly such as subscription or licencing. Specifically, we provide a "buy from amazon" link on each show details page. It is less intrusive than for example "we see you want movie x, buy from amazon now" on the profile page, and it will be interesting to see how this strategy pans out. We believe there is an intention to purchase, and that it will be a matter of finding the right time and location for that purchase within the applications processes and feature set.

Given the friends (social network) focus of the application, by far the largest suggestion we have received from our beta testers is to integrate into facebook. Not opposed to the idea, we have considered it, although at the moment we are concerned about the potential for an extreme increases in the size of the user base (if successful), without the confidence of clear monetization. We are bootstrapping these operations after all, and monster sysop costs without an active revenue stream may not be a productive place to be.

September is a new month, and we are actively pitching to each other regarding the next project. Some consider a facebook conversion as a viable project, taking Screen Sponge to the next level. Others wish to revisit Spicy Elephant, still confident that the niche of intelligent online study aids can be dominated.

We have lost Craig, (previously a full time developer) to contract work for the foreseeable future, and gained Matt, (our UI guy) full time, quitting his job to focus on our business ventures. We are also going forward without Cam, our forth developer who, although will still be kicking around, is changing direction to work on some ideas closer to his passions.

My position is that two seeds have been planted, that with ongoing maintenance may bloom into successful communities. I still yearn for my own business, and have proposed the direction of a B2B webapp targeted at the so-called fortune five million. I'm proposing for us to become an ISV for the long tail of small business, and to build an application with real (potentially paying) users on board from day one.

I want to hit harder, perhaps extend the prototype development time from one to two months, and focus more on happy users, aesthetics, and culture code. Given the tapping out of Craig for Matt, I expect the next project to have a much stronger user experience, and to be a hell of a lot more pretty.