Startup Founder Lessons: Fahd Butt on Why thinkpanda.com Failed

Fahd Butt was the founder, Chief Technology Officer at Thinkpanda (An academic research and collaboration platform). He is now the Founding Engineer (Employee #1) at Y Combinator-backed Nowmov (Nowmov is an endless stream of videos recommended to users based on their interests, friends and what’s trending around them)

So here’s my analysis of why my first startup didn’t get traction and why our team dissolved. A lot of these seem like obvious mistakes that I even knew about, but in the context of actually running a startup, sometimes its difficult to avoid certain decisions especially if no one is there to advise your team out of tunnel vision. Hopefully this will help an entrepreneur right now and will be a great resource for me when I decide to go on my own again.

Product

• Our product decisions weren’t grounded. We had a grand vision of disrupting the way knowledge was shared and consumed, especially in the context of research. In fact, we basically were building Google Wave (before knowing it existed in Google’d dungeons) and decided we needed to find a niche after we found out about it, cause they would kill us due to their real-time goodness (ironic now that we’re both dead).
• While the vision was great, we started building towards it, and this resulted in us spending a lot of effort in trying to become a platform. The “App Store for Academia” without a single killer app. In retrospect, we should have focused on a single use case that would inch us towards that vision, and not try to leap towards it.
• Due to the lack of focus, we spent significant effort building a lot of small features without any business/customer validation (RSS feed, integrated search) but because we thought it would eventually be useful.
• We spent a lot of time trying to understand how information is organized and what the best hierarchy is (groups within groups, and how aggregated information flows from one to another). This wasn’t a bad thing, but we spent way too much time based on some hypothetical edge cases of privacy in research and how researchers form groups within organizations. In the end, I decided to settle on the simplest solution and ignore the edge cases (coincidently, its pretty much the organization model Facebook Groups has right now)
• The Inbox Problem: We were building a product that competed with the stickiness and retention of email and we didn’t realize it. Email has proven to be the cockroach of online collaboration tools due to its flexibility and simplicity. Our organizational system, our notifications, our various integrated services, our app platform – none of these things mattered to our users who were comfortable and embedded in using emails to share and comunicate with their ad hoc groups and teams.
• I know I personally spent too much time thinking about scalability issues. But this was partially due to the fact we were on a shared hosting service (Media Temple) and weren’t strong backend engineers – so we did face some problems due to bad queries and database design. On the plus side I learned a lot about that part of the stack even though I started off as a front-ender/designer (like reading up on NoSQL debate for fun and the joys of de-normalized data). I just think it was a distraction having bad performance when all you want is to get the product out to users and see what they think (without the confounding effect of long page load times)
• Our startup pivoted from a university design project that was a different idea (social annotation tools like Diigo , Twine ), yet we stuck to a lot of the previous ideas mental models around information organization and to some extent, we were confused about what problem we were solving and what market we were addressing. Which leads me to…

Market

• We’re definitely not the only startup to have this problem, but this is what really killed us. For a majority of our 1.5 years as a startup, we didn’t really have a clue as to who our product was going to be used by. That’s not good.
• It felt that every so often, we would change who our product was for (is it journalists, researchers, students, professors, small businesses, other startups, consumer play!?). I think this contributed to the lack of product focus, because how do you know what to build to if you don’t know who for? So we started building around our own perception of solving an abstract problem – information overload for groups – what does that even really mean?
• Who were our competitors? Similar to not knowing our users, this also meant we didn’t know who our competitors were. It felt overwhelming because it felt that everyone was in our space: Google,
  Evernote, Yammer, Basecamp (web app), Academia.edu, Mendeley
• We did finally decide to focus on “academia” even though that was an option the whole time and we were sort of building towards that. We didn’t know too much to begin with about Academia but given we were recent graduates, had some contacts at the university and graduate level, we did as much as we could to talk to the right people. But Academia has a long cycle for projects and inserting yourself in the middle without any way for people to move their knowledge-base and the mental friction of a new unproven system by three undergrads is an uphill battle.
• It also felt that no one was prepared to make a decision. Professors would point us to admins, admins would point us to organizations, and no one had any money or motivation to change the status quo. We applied to several incubators and had online interviews – all had the same feedback: your market is not monetizable, its too slow, get out of it now.
• This led us to backtrack from the focus we had finally attained and to start looking at other markets to enter. We tried conferences, SMBs, startups, consumer forums (like Gravity or what is now Facebook Groups) but those weren’t serious efforts and there seemed to be much more developed solutions for all of those. I was strongly for going towards the consumer play and make the product like Facebook Groups, but this also meant that our whole product effort around building a platform had been wasted, it had nothing to do with our vision, our product was too complex for a simple forum and in the end is this something we can monetize soon (we were after all self-funded and had been going on for a year full-time).

Team

• I think we had a good break down of founder roles: biz, backend and frontend. However we worked remotely the whole time, and while me and the other dev would talk often about product problems, there was a greater lag to sync with the biz end of things. It wasn’t the fact that we were working remotely as a team, but that we didn’t project manage ourselves and didn’t have regular syncing meetings for more than just product decisions.
• We spent more time than we should have on trying to find extra help when it would have been better if we tried to simplify our problem so that 3 people could verify its validity.
• Our personalities were a bit too similar and this resulted in a quite a bit of group-think. There were a few instances of constructive disagreement, and generally I found myself on the odd end.
• One of the early decisions our group ended on, which, I did strongly argue against, was that of not being social. We were to be the anti-social network, a site that wasn’t focused on virality and making friends, but utility. Getting Things Done. I couldn’t convince my team that the two ideas weren’t mutually exclusive. I think this was the fundamental philosophical difference of the approach between me and my teammates. We ended up purposefully designing our site so that it wasn’t “social” (e.g. all groups created were by default secret/private). So there was no idea of activity happening on the site (this has since changed as you now see a global newsfeed of all public activity). Its one thing to avoid the use of the word “friends” in a professional/research context, but the underlying social graph is very powerful and that should be leveraged as much as possible to bootstrap an early community. As a collaboration/communication tool – this severely hurt us at the start. Our problem was that we were looking at users of Thinkpanda as isolated islands instead of building a community around sharing knowledge (which was our vision to begin with).

At the end we couldn’t agree on where to take Thinkpanda for its next pivot and decided to go on to move on to our next gigs. I don’t regret a moment of the time working on Thinkpanda and my founders have been solid teammates from the start till the end. I learned more going through this than sticking to the BigCo consulting job I quit from.

Personally, I felt I had a lot of gaps in my knowledge as an entrepreneur, a designer and an engineer. I wanted to learn with a team of strong experienced engineers and have fortunately found that at nowmov.com which I joined two months ago. They’ve got

• a fascinating product (people watch TV for ave. 4 hours but online videos for 15 min – how can we use curation and social dynamics to change that?)
• a competitive market (video consumption is greater than all other web traffic according to Wired, Youtube hasn’t innovated since Google bought them out, Apple/Google/Boxee/Roku are going to be great platforms to build a curated experience on)
• rock solid engineering team (couple Apple engineers who worked on Quicktime and iTunes, and another guy who has founded 2 startups) and great advisers and investors

This post originally appeared on Quora