People have been talking over the weekend about the failed Microsoft/Yahoo takeover talks. As a Yahoo shareholder and former employee I’m disappointed that it’s going to be quite some time before we get our money out of that stock (note to self: in the future, always sell ESPP shares the day you leave). I empathize with my many pals who work for both companies — however you feel about the takeover, it sucks having to go to work every day under a cloud of uncertainty.

But as someone who has also made his living on the Microsoft platform for years, I feel like I’m in a position to think about what Microsoft — the company that has been desperately casting about for a cogent internet strategy since 1995 — might do next.

When I was at the Web 2.0 Expo week before last, I was talking to a guy who asked me what I thought of the show. I told him I was disappointed — there were a lot of companies who were working on “me-too” social networking products, and a lot of old school companies that really had no business being there (Disney, Oracle). But I didn’t see anything really exciting or disruptive. I told the guy that I had yet to see the next company that was going to put 100,000 people out of work. What I meant was that I didn’t see anything industry-changing.

The ultimate type of business disruption — and the one that is ultimately the most difficult to achieve — is to disrupt yourself. I’ve said for a while that one of the reasons why Microsoft has failed to build a compelling network of web properties is because they are afraid of disrupting themselves. If you want to make a great leap, you can’t serve two masters — and in Microsoft’s case, they’re serving both the Windows master and the Office master. It’s pretty obvious that this is the case when you pay attention to Microsoft’s words and deeds — calling your maps product “Windows Live Local” probably isn’t the most egregious example, but to me it’s a clear sign that whoever named this thing isn’t really thinking about kicking ass, they’re worried about not upending the status quo.

Well, sorry guys. The web will continue to be about upending the status quo for some time to come. You can either play it safe and be upended, or be the one doing the upending. For companies like Microsoft and Yahoo!, there’s really no middle ground.

At the same time, many people in the past few years have said that the “nuclear winter” atmosphere for technology IPOs can’t last forever, that there should be some alternative for web startups short of an IPO. Some people have referred to this as the “rollup” — one well-heeled company that would go around developing products, or acquiring companies as needed, much in the way that Cisco used to do.

Microsoft should create a company to do this. A new company — an independent subsidiary — not beholden to anyone in Redmond. This company would have the mandate to acquire and create great web properties that people actually want to use and that will make money. No one inside Microsoft would have day-to-day control or authority over this company (aside from the sort of high-level oversight that you’d expect from a board of directors).

The new company would have to be based in Silicon Valley. It would have to reward its employees with equity that actually means something. The company must have the freedom to use the best technology to create the best solutions — and if that means Linux, or whatever, so be it. The new company should be permitted to do whatever it takes to attract talented workers from wherever it needs to — including poaching employees from Microsoft itself, if that made sense.

The new company would have an opportunity to throw off all kinds of bad habits, not only those practiced by Microsoft, but by large and small internet companies everywhere. Experiment with new ways of selling software online (not just licensing, and not necessarily advertising-driven). Put in the effort to make everything you do accessible from any mobile phone on the day it’s launched. Support stuff like OpenID and give users the freedom to add and move their data around freely. Dare to release a new product each month (as my team did when we were at Yahoo) and let the marketing and PR people catch up with you if they can. Reinvent web email in a way that makes Gmail look like a childish joke. Dare to sell products that will sell 1 million copies at $20 each instead of products that will sell 1,000 copies at $2,000 each. Provide standard APIs and interoperate with competitors’ products — not just as a way to kill them, but in a way that serves the ultimate interests of users.

Because that’s really what this is about — realizing that your customer is not the IT manager of a medium-sized midwestern insurance company, but the 120 employees of that company. Forgetting the user in favor of the IT manager (or your competitors) is why most users hate most enterprise software; I assert that it’s also why a lot of people have been frustrated by Microsoft today.

A big part of this would also be staffing the right people into the right roles. Microsoft’s geographical location (i.e., far away) and its historic resistance to hiring senior people from the outside (i.e., people with experience running large internet properties) act like a giant anti-talent force field; the new company would not have that problem. People would want to work for this company for the same reasons they wanted to work for Microsoft in the 80s and 90s — to get amazing things done and to have an impact.

My intuition has always been that Microsoft should have created this company ten years ago, and that they might have actually done so if they weren’t in the middle of some fairly profound legal problems. But Microsoft should not let its antitrust problems of the past prevent it from doing this. Just because the federal government tried to break up Microsoft in the past does not mean that spinning out an independent web subsidiary is a bad idea today.

Creating a new and independent web product organization could finally put Microsoft on the map with respect to the internet. It would take a few years, but it sure as heck wouldn’t cost them $45 billion.

Update: In an interview yesterday, Bill Gates said that Microsoft won’t be pursuing any new acquisitions to make up for not closing the Yahoo deal. He also said “now at this point, Microsoft is focused on its independent strategy.” Wait a minute: Microsoft has a strategy? When did this happen?