If you use OneNote, it’s likely that you read Chris Prately’s blog. As he has been called by Robert Scoble: “one of the geeks that runs the OneNote team.”
Recently, he had a good (actually great) post about the SP2 update to OneNote. That post is here. You should read it and the comments that follow. Go, I’ll wait.
There are a couple of fairly valuable lessons in this post for you and your start up. Start up of any size, by the way. This isn’t a Microsoft love/hate/they’ve changed post, so stay with me.
First, the meat of the post is a point by point review of each of the major errors/bugs/crashes that were fixed in SP2. Nothing surprising here, right? Well, I dunno about you, but having one of the designers tell me what they fixed, what didn’t get added, etc, tends to make the relationship I have with the product and by extension the people who made it, personal.
Lesson: Make it personal. Back in the good ol ‘compiled fresh daily’ days of shareware, freeware, etc, this was totally normal. The open source community, from what I can tell, does a good job of fix history, etc. Don’t forget this when you launch your product or service. The blog entry by Chris is a good template.
Next, you will notice a number of references to the Watson bucket. For all those, sorry, can we phone home and tell them we puked messages you’ve send, feel good that they are used and products are getting improved based on real pain points. This, to me, is net good even if there is a sizable group of people that worry about privacy, etc.
You will note in the comments somebody made the observation to the effect, wow, somebody reads error reports.
A nice fellow named Peter Engrav of the development team responded with this comment:
“It's actually better than that mere humans read the error reports. Every (reported) failure gets tallied on a server somewhere and (this is the slightly tricky bit) "bucketed" with other instances of the same failure. The first few instances in a bucket come with a chunk of extra data describing what precisely was going on when things blew up - thereafter we just count how many times the bucket is "hit". So we can work on fixing them in order starting with those that occur most frequently. It turns out the frequency curve tends to start out quite steep - a relatively small number of unique bugs cause a surprisingly large number of failures.”
Lesson for MS: Value add, folks. Making that data available to developers who are working on Windows applications, helps everybody. Great way to help developers build better applications. I believe there was a movement within Microsoft to get this data out to third parties. In other words, if your app blew up and Watson caught it, shipped the problem to MS, you could get this data. I don’t know what the status is on that project. Last time I asked, June, it was in the works.
Remember, in the end, customers just want value. Value for the time they invest with you as well as the money spent. By getting out in front of issues, keeping customers informed and, keeping it personal, you will build a long term successful business.
Yes, transparency is good... and those folks who actually want a better prodcut for free can always check out Evernote :-)
Posted by: Zoli Erdos | November 01, 2005 at 06:25
The information is indeed available to third party developers:
http://blogs.msdn.com/449865.aspx
Posted by: Raymond Chen | November 01, 2005 at 09:10