After Atlanta Startup Weekend I needed some time to recharge. A nice break from the Internet in general and blogging in particular over a Thanksgiving holiday that I
turned into a week’s vacation offered me a little time to reflect on
the lessons from the undertaking. As a facilitator more than a task
participant I have a unique view of what transpired and the takeaways.
Some of them are new, many just reinforced previous experiences with
startups. Here they are.
1. The Atlanta technology community is strong. We had over 120 people sign up for the weekend. We also ended up with several firsts. Jeff Haynie walked in with a prototype on 8:00am Saturday morning. Sanjay Parekh and Michael Mealing put a provisional patent together. On top of those a Google search for the brand name of the service we created returns
more results than any other Startup Weekend product. And, most
importantly, we launched.
2. Good ideas are immediately apparent. When Paul Stamatiou first suggested the idea that became Skribit on the idea wiki, you knew it was a winner. It made it to the final four (but was rated lowest of the four.) Once it get that far it took a little lobbying. : )
3. Quick decisions are needed. Andrew Hyde has some amazing
facilitation skills. While I was pretty involved in the planning of
Atlanta Startup Weekend, one of my biggest worries was how we were
going to go from over 20 ideas, down to 1 in less then five hours.
Andrew led us through this process masterfully.
4. Make something that people want and can grow organically.
We ended up selecting the Skribit concept for a lot of different
reasons. One of the bigger considerations was that the group felt the application had the highest viral element to it when compared to the other candidates. Skribit has the ability to grow by itself without a sustained business development or marketing effort. The early traffic and use numbers support this belief.
5. Somebody has to be in charge of product. When Andrew
first broke us into groups Erika Brookes mentioned that it seemed odd
that there was no product group. Did that one ever come back to bite
us. Business development, marketing, and user experience were all
telling dev what to build. Of course dev had their own ideas as well.
6. Leaders step up and lead. Alan Pinstein was the second person that brought up the product lead issue. Which led to a product development intervention on late Saturday afternoon. During this meeting, while many of the said groups denied setting product requirements, Jason White stepped up for the first time with a solution on the dev side. During a user
experience meeting about an hour later a marketing person was
introducing a new feature. Lots of folks were indeed setting
requirements. Shortly thereafter Alan became the point person on
product requirements and Jason took charge of the development effort.
A bit later in the evening I was accused of being leaderish.
7. Get the product out. Late on Saturday evening it became
pretty obvious that we were not going to launch if we continued down
the same path that we had been following. The Appcelerator
team was confident that their platform could pull us out of the weeds.
The decision was made to let them have a go at it. They
delivered. It was amazing to look over Jeff Haynie’s shoulder while
Alan P was diagramming the product requirements on the white board and
then have Jeff bring quickly bring the product to life on his
platform. By the time they left that evening most of the application
modules were working independently of each other.
8. All startups need milestones. Driving back to ATDC early Sunday
morning it dawned on me what we had been missing thus far in the
weekend. Milestones. All startups need them. We had none. Quick early
huddles with Alan and Jason led to setting milestones of noon for an alpha
and 7:00 pm for the first beta. Not sure we made either, but having
the goals made us focus and we reached our ultimate launch goal.
9. Development has to be in sync with compartmentalized tasks.
To get in a position to win we had to take a short cut. The
development team was not in sync Sunday on the product requirements.
With the rate of code commits this created a situation where the devs
were stomping on each others code. This caused hours upon hours of
rework, some raised voices (including mine) and nearly led to us not
getting the product out.
10. Marketing needs to follow product. Unfortunately the
marketing team was not in communication with product development as
well as they could have been. Around 9:00 pm on Sunday they contacted TechCrunch and Guy Kawasaki. They did an amazing job. Both of them covered us. Which led to a good deal of traffic. To a site that had nothing but a logo on it. No
app. No way to capture email addresses. A mad scramble ensured to
correct the latter. We had 6,000 unique visitors within one day of
launching. If you Google ‘skribit’ you will find over 72,000 results
on a term that did not exist two weeks ago. No pr is bad pr.
11. You have to believe in yourself and your team. At 11:45
pm Andrew asked if we were going to try and launch by midnight. After
about seven minutes of intense discussion the team decided yes. Nate Clark
frantically scrambled to move the code from the dev to the live
server. We were a launch. We stayed for about three more hours fixing
things as they broke, but we were a launch.
12. Engage experts. At some point on Sunday, Andrew said
something to the effect of “I could have told you this was going to
happen” to which I replied “why didn’t you?”. “Nobody asked me” he
said. Andrew was and remains the smartest guy in the world about
building a startup in a weekend. We should have been more proactive in
reaching out to him. As all startups should be with their expert
friends.
While there were some rumblings here and there, I think in the end
Atlanta Startup Weekend was a smashing success. The only question is
how long should we wait before we have the second one?