Items Tag › matthart

As I discussed in an earlier post, echominder.com is shutting down, and the time has arrived. March 1st was set to be the final date, but since all activity has been completed (e.g. all scheduled calls have gone out), I’ve gone ahead and shut down a few days early. Here’s a short walk through the evolution of echominder: Summer 2005: MIT Intern Bill W along with Anna run the DAIQURI project - Digital Assistant Integrated Query Utility for the Immediate Retrieval of Information … we give our interns a lot of leeway…. That project went through a number of prototypes and tests, morphing into two primary needs, one of them being “note taking on the go” - the basis for echominder. Now it’s more than 6 months until Matt (me) picks up the idea and builds a prototype using VB, Skype and Gizmo. It’s not very good at recognizing the callback time or doing any kind of transcription, but it does great at recording your message and forwarding it to your email. That might not sound like much today, but it was a year ahead of any free service doing that (that I know of). Intern time again, summer 2006. This time MIT Intern Anunaya P and I start looking at using the recently released Asterisk open source PBX system. We get it connected up to Gizmo, then to VoicePulse, the service we eventually use for echominder. The hardest part was just getting the hosting setup, with a new DMZ isolation from the Intuit intranet and ports opened for SIP communications. But it still wasn’t echominder - not yet. It took another long period before we finally, formally, adopted it as a real project. The problem was that we weren’t a development team, rather more of a research team. Sure, we had developers, but it wasn’t our purpose. We couldn’t get a business unit to take on the project, so we finally started building it ourselves. Scott took on the website, I built the IVR/Asterisk back end, and Holly managed the project. We finally released in the fall of 2007, after a half day of disputing whether or not we were ready to launch. Part of that dispute was due to the fact that, well, we were really late. Slow from idea to launch, and competitors had begun to appear. How would we differentiate our service? Were we top-notch within that difference? Well… no… not really… but we launched anyway. Here’s a few tidbits we learned/experienced along the way to launch, and shortly thereafter: We first named it “voomerang” - Arno came up with that one - but it was taken. If we’d have developed the idea as soon as we came up with it, we could have been a year+ ahead of anyone else - so hurry up and build/launch!! We heard from customers that “Incorrect transcription is okay as long as you are getting better.” In practice, transcription needs to be nearly perfect for it to be acceptable - and don’t ask stupid questions like, “Do you want transcription?” Duh… of course you do. It’s the first number on Anna’s favorites list Jonathan and I will need new ways to capture thoughts on the road. The “ultimate” phone reminder service needs to have accurate transcription, voice recognition, and a simple-to-use website. (Scott) Most users used the service for their own reminders vs group broadcasting. (Scott) And one of the biggest things we learned: if you can’t dedicate time to it, shut it down… no matter how much you like it. Holly got pulled off to manage other projects, our charter changed, I’m doing a dozen things at once, Scott is building Intuit Labs… So I jumped in with the full knowledge that my primary job would probably be to shut it down. We had new features developed, such as voice recognition IVR, unlimited schedule-ahead, better groups… but ultimately, we didn’t have the bandwidth to support it. We shopped around the BUs again, but they couldn’t take it either. With competitors coming of age, we saw our future roadmap getting developed by other folks. So most of my work and planning as the PM of echominder consisted of managing the sunset decision and process. We were deliberate, with a DACI (similar to a RACI) of folks chiming in, but in the end, it came down to focus - something we couldn’t spare for echominder. Fortunately, we are keeping the technology in place. I’ve developed several ways to leverage its capabilities, and other applications are starting to use it. So while the product has gone away, the technology has remained and will be powering other mobile applications at Intuit. So long echominder.

matthart

echominder has left the building

echominder is shutting down on March 1st, 2009

Popular Tags

Search