A Debate With Microsoft On The Departure Of Don Dodge

By Dharmesh Shah on November 22, 2009

Here is a hypothetical conversation that might have happened (but didn't) with some high falutin’ person at Microsoft on the whole Don Dodge thing. 

Imagine a scene where I’m sitting in a Seattle coffee shop having a clandestine meeting with the Microsoft individual that was involved in the recent, highly publicized departure of Don Dodge.  (If you don’t know anything about this topic, you can get caught up with this TechCrunch article and it’s 200+ comments).

Dharmesh:  How could you allow Don Dodge to quit Microsoft?  Seems someone wasn’t working hard enough to keep him happy.

Concerned High-level Microsoft Person (CHMP):  Well, um, he actually didn’t quit.  We let him go…

Me: Choking on my decaffeinated coffee (which turns out does actually contain some caffeine, enough to keep me up at night, which I don’t really need any help with).  What!?  Why the heck would you do that?

CHMP:  Well, it was part of a much larger reduction in workforce event.

Me:  Um, ok, but weren’t there tens of thousands of Microsofties that weren’t let go?  Why wasn’t Don one of those people?  Seems if things were completely random, there’d be a higher probability that he’d wind up in that bucket instead of the “let go” bucket.

CHMP:  Well, it wasn’t really random.  It’s not like we pulled names out of a hat.  At Microsoft, random is a 6 letter word. In any case, we actually put him on the list to let go.

Me:  Oh, I see, so you’re really saying that the whole notion of attracting developers onto the Microsoft platform — particularly getting startups on board while they’re still young and impressionable is no longer a strategic priority. 

CHMP:  Well, I wouldn’t quite put it that way.  We’re still all about “developers, developers, developers”.

Me:  And you don’t think that Don was helping out enough in terms of putting a semi-tolerable face on Microsoft?  Do you have a sense for how much discussion has been going on in the blogosphere now talking about how stupid this move was?

CHMP: Well, in our defense, all of the recent fall-out from his “departure” happened after the event.  We couldn’t possibly have known that Michael Arrington and a bunch of other high-influence people would react so negatively.  We figured, hey, we’re letting a bunch of people go.  This is one of them.  It’s a bad economy and all that…

Me:  So, you didn’t consider the PR impact at a time when you can use all the positive PR you can get?  You didn’t know that Don had a blog, or that you paid him to fly around the country and evangelize and build these strong connections in the tech industry?

CHMP:  As I said, we couldn’t have known that the response was going to be so negative. 

Me:  And, what about the fact that he got hired by Google.  And that he’s now switching to Google products — and writing about it.  This can’t be sitting well with you.

CHMP:  Well, yeah, we’d prefer that he had not gone over to one of our arch rivals.  In fact, Google is the last place we wanted him to wind up.  And in terms of his switching to Google technology, he’s just one user, we’re solving for the “long term”.

Me:  So, let me get this right:  You had him on the list to let go, you knew he had reasonably strong brand and a community of followers (including some high-level folks) and you knew that Google was high on the list of places that would likely recruit him. 

CHMP:  Yes, that’s kind of right.  But, we didn’t KNOW that the PR cost was going to be so high. 

Me:  And what about the whole Boston thing?  Wasn’t it kind of nice to have a well respected, likable guy that was singing Microsoft’s praises in the second-largest tech startup area in the world?  Don’t you think it’s going to hurt your “long term plans” to lose some of that mojo?

CHMP:  Ok, between you and me, we were idiots.  In the whole grand scheme of things, Don wasn’t that much money.  A mere rounding error in the PR and evangelism budget.  And it’s not like there were 50 people like him running around at Microsoft.  He had some unique value, some great connections and did an exceptional job evangelizing Microsoft while still maintaining credibility (and dignity).  That’s not easy to do.  And yes, we knew there was the risk Google would recruit him — they’ve been known to do some smart things.  And yes, we’re kicking ourselves now because from a business perspective, Don was accretive.  But, let’s talk about the real reason this all happened.  Given infinite time, we could have likely figured out that this was a bad decision.  But, realize that we’re Microsoft.  We’re big.  We can’t afford to analyze each individual decision like this.  Thing of it as the price one pays for selective focus.  You can fault us all you want, but hindsight is 20/20 and foresight is expensive.

[end of scene]


This brings me to the startup moral of the story.  You should be investing considerably in your team.  If someone is profitable (creating more value for the company than they’re costing you), then there’s only reason to ever let them go:  Their “payback” period is longer than your cash resources will allow.  Said differently, they’re great, and they’re going to make you money, but you can’t afford to keep them.  That’s a legitimate (and often common) situation. 

You should also factor in the the likely outcome of letting someone go.  Think through the likely scenarios.  One additional high-level question I often ask myself that is looking to leave:  Would I rather this individual work for me, or my strongest competitor?  If I want them on my team, I fight hard to keep them. 

What do you think about assessing the value of members of your team?  Do startups tend to over-estimate the competitive risk of an employee leaving?  Do we make mistakes in terms of choosing who to let go when it’s necessary to do so?  What do you think?

Continue Reading

Answers OnStartups Roundup 1 - March of the parentheticals

By Jason Cohen on November 20, 2009

By now you've probably heard of Answers OnStartups, the new Q&A community for any topic about startups and entrepreneurship. If not, here's Dharmesh's announcement.

I have the good fortuneonstartups questions answers of being asked to co-moderate the site with Dharmesh, probably as a result of the popularity of a few provocative but useful guest-posts here on OnStartups. (Note to those of you hoping to expand your influence in the blogosphere and beyond -- take the advice from Leo and Darren and Mary and, I suppose, me: Guest-posting is a good way to do it!)

What does "moderation" mean, you ask? I took it mean "Get the CSS in order, make sure there are no JavaScript errors, and correct everyone's spelling errors. Oh, and be the #1 highest reputation person on the site." Durn, that last one actually took some effort... (Yes I see you gaining on me Alex!)

But, you know, it's not a competition. Yeah, right. That's why there's voting and reputation, 'cause it's not a competition. Riiight.

As some of you know, there was an unexpected problem with the site at the beginning. See, Answers OnStartups is built on the same technology as StackOverflow, through Joel Spolsky's StackExchange program. (Yup that's our Dharmesh on the front page! Which reminds me, that's a good lesson in marketing. Q: How do you get your name and URL on the front page of a high-traffic, highly regarded, high-SEO website like StackExchange? A: Write a fantastic (but honest) review of the product and give it to the owners. Make it so awesome that nothing they could say about themselves would be as good or believable as your words. Then they'd be foolish not to!)

(And since I'm on the subject, and since I've already interrupted myself with this many parentheticals, you might be interested to learn that, technically, StackExchange is FogCreek but StackOverflow is Jeff Atwood.)

Anyway. Where was I? Oh yeah, problems bootstrapping the site.

See, Answers OnStartups has an automatic permissions system where you have to have a reputation of at least 15 to vote, and 50 to leave a comment. Even more for things like inventing new tags. It's smart because it prevents people from (too easily) gaming the system by creating accounts and tearing up the place anonymously.

That works well at StackOverflow because they get a million hits per day (literally), so if you start contributing properly you get up-voted fast (sometimes in a matter of minutes). Very quickly an earnest contributor is granted these permissions.

But at Answers we had a bootstrapping problem. All these kind and wonderful folks came to use the site... and essentially couldn't do anything. The only people with rights to up-vote were Dharmesh and me (because we are blessed moderators, pronounced "bless-head"), and even then we were limited to 30 per day.

So all the initial members had to run around and up-vote each other for about a week, and newbies would come in an invariably ask questions about why no one is voting, which, ironically and usefully, got them a lot of votes.

The lessons from this little mishap:

  1. When you use pre-packaged software, you get the limitations along with the upsides. Just remember though, had you rolled your own you would have a different class of bug, probably worse.
  2. It didn't matter in the end, because what we're really doing is building a friendly community for startup discussions, and little problems like that aren't going to matter assuming you have intelligent members. And if you don't have intelligent members it's not going to work anyway. In other words, don't sweat the small stuff.
  3. Had we built our own, others would have beat us to the punch, and we would have lost the war.

As proof of that last point, there's now another StackExchange-based startups Q&A site! I'm not going to link to it because I don't want to give them the credit -- which I admit is kind of crappy of me, but in my defense I am writing in the throes of passion and cannot be responsible for my actions. Not buying it? Fine, but I'm linking to our question about it.

Anyway Answers OnStartups is better because we have a totally killer user base. Active users you've probably heard of include:

(I linked to their Answers user profiles too so you can see they're actually active, not just name-dropping.)

(I love the phrase "name-dropping." It sounds like you're taking a dump on the conversation which, often, you are.)

But perhaps even more fun is meeting all sorts of folks who aren't already famous but who are running interesting startups and have terrific, new things to add to the conversation.

Over the next few weeks, watch this space for highlights of these users, as well as highlights from some of the more action-packed discussions that are going on.

To whet your appetite, one of the most popular questions is: How do you pick a platform for a Web 2.0 startup company?

See that link for the full discussions, but here are some highlights:

  • The platform is not as important as other factors. There are hugely successful companies based on any technology you can name: Java, .NET, PHP, Ruby, Python, Perl.
  • A major factor is existing competence. If you know a platform/framework well already, that trumps other considerations because it means you can get started fastest. Speed to beta is part of success.
  • If you have no existing competence, it's most important to find/hire the best possible developers. Because it's always hard to find awesome talent, whatever platform they love is the right choice.
  • It depends on whether you need to get something running ASAP or whether long-term maintainability is necessary right from the start. If the former, Ruby and PHP generally get applications going fastest. If the latter, Java and C#/.NET generally have the best tools for that sort of thing.
  • Both Sun Startup Essentials and Microsoft BizSpark make certain application stacks particularly attractive. If you like the other benefits afforded by those programs, either are a good choice.
  • Pick a framework that will stick around for the long haul. Ruby on Rails isn't going to be ditched. ASP.NET won't be ditched but Microsoft is a moving target. PHP itself is stoic but any particular framework is more questionable. Which community do you think has longevity?
  • If your startup fails, which technology will leave you best positioned for future projects?

As with many questions on the site, there's usually not one right answer. As it should be.

So join the conversation! Come visit Answers.OnStartups.com and see what all the fuss is about. It's fun.

Continue Reading

Recommended For You


Let's Connect

And, you can find me on Google+, Twitter, and Linkedin.

Recent Posts

Chat with GrowthBot

It's a bot to help you with your marketing and grow. You can research your competitors, improve your SEO and a lot more. http:/GrowthBot.org