I think Cakewalk's demise was simply down to money and management decisions. Sonar is a great DAW, I don't think there was any fault in Sonar itself to cause Cakewalk's demise.

In any product business, sooner or later you have to manage the balance between new sales and renewals. Once you get to the point where you've saturated the market (or your market share obviously isn't increasing), then setting yourself up to comfortably survive on renewals should be the goal. That way new sales are a bonus, but you don't have to rely on them.

Of course keeping your existing customer base, and ensuring renewal costs are at the right level is vital here.

Here are a few areas where I think is where things went wrong for Cakewalk:

1. Sonar X1: The workflow changed significantly here, and although it was arguably better, it took me a while to get used to the new interface and workflow. It was also pretty buggy. Not everyone would have stuck with Sonar at this point, i.e. if you're going to have to learn a new workflow anyhow, then why not try out other DAW's?

2. The "rent to buy" / "subscription" thing: I think the confusion and initial backlash over this put a lot of people off and hurt Cakewalk.

3. Lifetime upgrades: This was a novel idea, but it could only be successful if combined with decent new sales and plenty of paid-for additional content. This obviously didn't happen. You need sustainable income from somewhere: either it's from upgrades, or it's from new sales or content.

So the alternative would have been to stick with the old model, however looking at the figures around the time Roland took over Cakewalk, this obviously wasn't working either.

I think all of the above was done in good faith to try to turn things around, but it just didn't work out.

As a software development manager myself who has just semi-retired a product, there's a few things that maybe Cakewalk could have done:

1. Recognise that Sonar is a pretty complete DAW, and focus on bug fixes/stability rather than new features for the core product... at least for a a year or two in any case.

2. Focus on ease of maintenance by investing in automation, code refactoring and removing any deprecated code that is causing maintenance headaches. In my day job we managed to get a 130+ workforce working on one product down to around 6 people (with most of the other staff now working on new products). It took about a year to get it to that state, but now at least it's financially viable in maintenance mode - i.e. the support revenue more than covers the remaining staff costs.

3. For new development, maybe focus on added value than on the core product. These could be modularised and sold accordingly. Things like:
 - Integration with notation software, wave editing software, composition tools etc
 - Hardware specific modules (like say having the ProChannel leverage the internal DSP's of RME / Tascam interfaces) - SoftTube Console 1 is a good example, but it was way too expensive.
 - Better project migration/import/export from other DAWS. IMHO this would be a HUGE selling point for pro studios (especially ProTools import/export).

4. For future core development, take a step back from trying to innovate. Let the other DAWs innovate first, try it out, and improve on it. It seems to me that's what Cakewalks's competition have been doing for years, so try beating them at their own game.

The most important thing here however is knowing your customer base, balancing cost/return on investment, and getting your priorities right. There's no point in introducing features that won't be widely used or appreciated.

Just my 2p's worth...

M.