Some people see Microsoft as a company in decline, in part because it draws such a large part of its revenues from old and not that innovative business lines. (It does, however, still make some innovative products.) But reputations aside, what of its real world effect on startup culture?

According to people I spoke to from Seattle’s tech industry, Microsoft’s influence on the startup ecosystem has little to do with product or innovation inside the company and more to do with sheer scale. With 100,000 employees and billions in the bank, there’s always going to be some collateral productivity.

The phenomenon is true for Seattle as much as it is for the rest of the world. Because of its stature, and despite management missteps (like literally betting on a losing horse), the company attracts a wealth of talent, which can have positive spin-off effects for the tech community.

The talent draw can be good for Microsoft, but too often it only helps the company for a short time. Initial employee enthusiasm eventually dies away, and the large salaries Microsoft pays only keeps the entrepreneurial people at the company for so long. As a number of people told me in informal discussions, the bureaucracy and management infighting scares away too many good people.

Chris DeVore, a partner at the accelerator Founder’s Co-op, says Microsoft’s influence on talent can be negative because many of the staff who stick around are the ones who are best at dealing with internal politics, which in a startup can do more harm than good. “People that come out of Microsoft have developed a different set of skills – people and skills that aren’t necessarily good for a startup.”

The company’s ability to attract great people can also have spin-off effects in unexpected ways. As one founder told me, asking that his name not to be used in order to avoid potential legal issues, “Microsoft is one giant relocation fund” for Seattle. This founder worked for Microsoft for a grand total of nine months and moved to the city on the company’s dime. He suffered through his pain at the company until his contract allowed him to leave without repaying the relocation expenses. According to other people I talked to in the area, he’s not the first to use that strategy.

Microsoft’s large-company talent magnetism is probably one of Seattle’s most important overall strengths. Because of it, Facebook, Zynga, and Google have all opened satellite offices in the city. They’re there largely to steal talent from Microsoft and Amazon.

To be fair to Microsoft, it has been trying recently to be more involved in the startup community. It works with TechStars in Seattle, runs BizSpark, hosts Azure, and is persistent in trying to get its message out to the world.

There have also been a few great, or at least very successful, companies to come out of Redmond. Valve was started by ex-Microsoft employees, as were Zillow,, and Real Networks. Expedia is a major Microsoft spin-off.

However, with 100,000 employees, offshoots and outreach programs are nothing out of the ordinary. Compare Microsoft, for instance, to Apple and Google. Apple has seen offshoots like Android, Pandora, and Posterous, while Google’s resources and support have helped bring Twitter, Instagram, and Foursquare to life (and that’s to say nothing of Google Ventures).

While offshoots and byproducts are great for the startup ecosystem in Seattle and will boost the salaries of engineers, they don’t really help Microsoft. Yes, there is the “rising tide lifts all boats” argument, but in this case the returns are scant. Microsoft could do more for the community in which it resides beyond just being a leak for talent.

To get started, Microsoft should promote its most entrepreneurial employees. It should get more visibly involved with the Seattle community. And maybe it should even take a leaf out of Google’s playbook when it comes to investing in startups.

It’s telling that that, in recent years, the most noteworthy impact Microsoft has had on the startup community was a strategic investment in Facebook. Microsoft is important enough that its impact should be felt more often than once every ten years.