For a long time I was being secretive about what I was up to. I would tell people that I “had a couple of projects I was working on” or something vague like that. But the truth is I have been very busy working on exactly one thing since about the beginning of December.
A short while after I was let go, I was out in Chicago to see my old friend (and former CS STARS CEO) Bob Petrie. I wanted to catch up and bounce some ideas around, but he had a very specific idea to pitch. Bob had no intention of competing against CS STARS when he was first let go. There were too many friends still there. But after Jeff Markowitz got let go and then the entire NY office (some may not realize, but Bob started the business in NY) Bob’s thinking shifted. Not that we don’t still have friends at CS STARS, we have lots, but it has clearly changed.
Frankly, my first thought after being let go was not, “gee, how can I stay in the insurance risk management industry?”. But Bob’s pitch was compelling. Let’s build a RMIS system that has a subset of the STARS feature set, but the really important US RMIS features. Do them as well or better. Make the implementation/service model simpler and therefore cheaper. Carve out a decent share of the market. Own the business. Make some money. Work for yourself.
So I thought about it for a while. The job market is lame. I have severance pay. But this is a serious v1 software play. Bob’s initial thought was the two of us and some cheap development labor. I realized we needed more than that. We needed partners who knew what they were doing.
As part of my severance deal, I can’t solicit CS STARS employees. But the entire NY office was no longer employed by CS STARS. Including two of the best and two good friends. My pitch back to Bob was simple. Let me talk to Linus and Tim.
It turned out that both were interested. The folks at 37signals have stated that three developers is the optimal number for v1. In this case I think they were absolutely right. But you have to have the right three people.
Linus, Tim and I have probably worked harder in the past four months than ever before. But working together (with Bob, while he isn’t programming, he’s a big part of the project) has been incredibly easy. We all know what we need to do. We discuss, we take on the parts we want to, we go off and we work. We get back together, discuss, rinse, repeat. There is 0% administrative non productive time. Boring conference calls? Stupid paperwork? Mindless time tracking? Non existent.
And after a relatively short time, we realized that we could build something pretty cool. Not just something that was ok and cheaper. Linus started with the name Origami and I really liked it because to me it evoked elegant simplicity. That has become our design philosophy.
It’s the 37signals philosophy.
We believe most software is too complex. Too many features, too many promises. Instead, we build simpler web-based software with elegant interfaces and thoughtful features you’ll actually use.
And the classic:
it’s better to make version 1 half a product than a half-assed product
Origami doesn’t do everything. But what it does, it does well. And that is much more fun to work on.
We had the rare opportunity to start from scratch. No legacy issues. SaaS only. Single database. Data model from the ground up with everything we know about what the product needs to do. That last point has been unbelievably huge. Can you imagine running a loss triangle on 2 million claims in less than 30 seconds with no pre-processed data? What else could you do with no legacy constraints?
Who knows where this will take us. At the end of the day, we have to sell it. Can we? We’ll see. But we have a compelling product, a great (if small) team, and we’ve had a lot of fun…
The next chapter has begun…