Subscribe by email

Q&A with Hal Higdon

Each week, coach and author, Hal Higdon answers your questions about running. Here's the latest:

« Coaching Tips: Know Your Athletes | Main | Interval Workouts for Triathlon »

Gear's Thoughts: One Year Post-'Classic Mode'

Gear FisherThis post is part of a new series by Peaksware CEO and Co-founder, Gear Fisher.  He's been a driving force behind our innovation since it was just an idea thrown out by a couple of frustrated coaches over a beer more than a decade ago.  Each month he gives his thoughts about what's going on inside the company and where we're headed.  This month he takes a look back at one of the biggest challenges our company has faced.

It’s hard for me to believe, but it’s been just over a year since we migrated entirely off our ‘classic’ code base to the current version of  If you’ve been a customer for more than a year, (and many of you have been with us for 7+ years), you’ll certainly remember the transition and the challenges it presented.  Lets just say, it was a rough time for all of us.  Lots of stress, worries and frustration to say the least.trainingpeaks classic mode calendar

However, we learned a lot from the process. I spent much of my time coordinating with our product development team on what features to build, in what order, and with many of you on personal calls, demos and webinars. Suffice to say, change is hard, especially when it’s on someone else’s timeline.

For those newer customers, and for a quick background, the “big migration” started for us in 2006. We had been bolting on features to our software for 6 years, the code was getting out of hand, and our technology had lost pace with the fast changing web world.  We could see our growth slowing too, we knew we had to re-engineer our platform to make it more scalable, modern and supportable for the long term.  We also knew, that for the first time in our history, we needed to hire developers and spend money ahead of the curve, which was money that we didn’t have.

At the time, we were financed 100% by organic growth, we had no debt, and we cash-flowed the business monthly, we were small, but we had some big dreams.  To accomplish those dreams, we sought an investor that would provide us the capital to rebuild our infrastructure and grow. We have never mentioned this publicly before, but after a year of looking, we were incredibly fortunate to find a customer that was passionate about what we do, and we raised a significant amount of capital in late 2007.  This gave us the ability to hire exceptional talent and begin the development of “TrainingPeaks 3.0”.   We started the rebuilding in earnest in January 2008.

trainingpeaks classic mode dashboardIt was risky, taking thousands of people (you guys reading this!)  and convincing them that we had ‘a better way’.  On top of that, we’d be on the hook for the first time ever with ‘outside money’ and frankly, it was pretty scary.  We couldn’t just wall-up and put ourselves in a closet for a year, we had a real business to run, with thousands of real customers, so we knew the effort would have to be done in parallel with our existing business demands.  It was difficult to manage, but we had some incredible talent on our team.

After about 8 months of development, we started to leak out our new system’s features, you can see that by scrolling to the bottom of our 'What's New' page. We would basically roll-out some very early concepts, get public feedback, and tweak the system. We had to port thousands of features and millions of lines of code from our classic system to a new one.  Not only did we rebuild our user interface, we also rebuilt the layers in between which encapsulated all of our business logic.  It was a huge effort.

After almost two years of hard effort, we’d gone through many alpha and beta versions, updates and changes.  The new platform really started to gain traction in December 2009 once we ported over the VirtualCoach.  We then set a cutoff date where we would end-of-life the ‘classic’ system and rely entirely on the new system: Feb 17, 2010.  Once customers realized there was no going back, we were skewered by a lot of people.  The Slowtwitch forum predicted our immediate demise and many coaches and athletes alike screamed bloody murder, in public and in private to us directly.  It was rough, but we were confident of our direction.

trainingpeaks calendarFast forward to today, a little more than 1 year since that launch.  We grew 49% YoY in 2010, it ended up being our best year in the company’s history. Traffic was up, we were profitable again, and our customers voted with their wallets, and we had done it.  All while the recession was taking place!  For that success, I am tremendously proud and grateful to you for gutting it out.  I’m proud of our team here in the office, and thankful to you for sticking by us in difficult times.  I’ve heard from many of you that you were admittedly reluctant and worried, but now you would never go back to the old system.  It’s music to my ears and confirms what I believed all along.

One thing we are good at, is iterating on a product.  And we learned a lot during that time in how to handle major change.  Going forward, we’ll continue to take your feedback, filter it, and improve our software to better meet your needs. Today, we’re working hard on mobile offerings and ensuring that we have the right software for the right screen, whether it’s your 24” monitor at your office, a laptop, your iPhone/Droid or a tablet, we’ve got a solution and will keep working to fulfill our promise of providing you with world-class software for endurance athletes and coaches.

Thanks for your continued support!


* * * * *


Did you know?  Gear's commitment to listening to customers and love for all things tech pours over into social media.  You can find him posting regularly on the TrainingPeaks Facebook page, answering questions via twitter (follow him @gearfisher) and spending some quality time in the TrainingPeaks forums.  Got questions or suggestions?  Please leave a comment below or contact us.

Reader Comments (4)

How about offering the old "Classic" version as Training Peaks-Lite at a discounted price? I still like the "quick-in quick-out" non-Adobe interface.
Or maybe just a spreadsheet? something I can open on a lap-top at a race, and down-load my race data. I want a low memory-intensive interface to grab my .csv files? call it KISS (Keep it simple stupid?)

March 31, 2011 | Unregistered CommenterJames Adams

Well...that throws new light on it...I bet you thousands of customers [free an paid] did't know how close to the wind you were sailing looked fine from hear. From reading this thanks for your honesty and integrity...I admit I was one of the whingers...and still like the snappines of the old Classic and less download cost for it to render and operate could it work on the mobile platform?
So you have given us a look back can you tell us what an whens in store for 4.0? Thanks keep up the great work

April 1, 2011 | Unregistered CommenterMilt Mac Aust

@James you can simply use Device Agent for downloading workouts offline from any device. Otherwise, use our free mobile app for iPhone (Droid coming soon) or the mobile web-app Those are very easy and fast HTML based interfaces that are very KISS.

April 4, 2011 | Unregistered CommenterGear Fisher

@Milt we have alternative interfaces (see previous post reply to James) that you can use too. We're pouring a lot of effort into expanding our online functionality and our mobile offerings too. There would be no snappiness of the old system given how many more people are using the tools now. LIkewise, the site should load in 5-10s after the intial load of the flex framework and libraries. Subsequent logins are very quick. But we'll keep refining and improving the speed. As Google has said, speed is always the #1 feature.

April 4, 2011 | Unregistered CommenterGear Fisher

PostPost a New Comment

Enter your information below to add a new comment.
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>