[raganwald]

I agree, let's flip this around and think like customers. We go out to choose a vendor for some important service.

Vendors A and B have roughly equal parity on features and services, but vendor A's employees eat their own dogfood and vendor B's don't.

Now in one sense, who cares? Eating your own dogfood is a means to an end, not the end itself, so it's like finding out that McDonalds employees don't eat McDonalds food. As long as they wash their hands, who cares what they eat themselves?

But on the other hand, I'm a human being, and I'm personally a lot more comfortable doing business with a company that seems to care about its product from top to bottom, and isn't staffed with people who don't like their own product enough to use it.

[jrochkind1]

Yes, employees that use their own product generally a great and encouraging sign.

But what if you find out that the employees at company A use their own product only because the CEO ordered them to do it or get fired, and they actually hate the product themselves too?

No longer quite so encouraging.

They are mistaking the indicator for the thing indicated. Dogfooding is an indicator of quality and commitment when it happens naturally; when you artificially compel the indicator, it's no longer a good indicator.

[raganwald]

They are mistaking the indicator for the thing indicated.

This is one of the most important (and neglected) insights in business. People often optimize for the indicator at the expense of the thing it is supposed to indicate.

Example: One metric is of customer dissatisfaction is unsubscribes. If you make your unsubscribe process difficult, unsubscribes go down. In reality, dissatisfaction might be climbing through the roof, and a difficult unsubscribe process may actually make it worse, not better.

That being said, there is a difference between: "If you don't care, go work somewhere else so that the only employees left here are the ones that care," and, "Do it even if you don't care or else I'll fire you."

The former might be about getting rid of unmotivated dead wood. Some people are good no matter what, leave them alone. Some are terrible for your company no matter what, fire them or entice them to quit. The remainder are the ones to manage.

But then again... You don't want to fire or push out the people who might be able to tell you that the dogs hate the dogfood because it tastes like shit. Which was the punchline of the joke that the entire "eat your own dog food" expression is based on.

[bigtunacan]

I think you are missing the point of the "eat your own dog food" mantra that originated from Joel Spolsky. No one wants to literally eat dog food as we all imagine it would taste terrible. The same holds true for your product. That product you have been working on for your company, whoever it may be, is total, utter, garbage, and if you have to eat it everyday you will realize it tastes terrible.

The more you use the product, the more aware you become that your product is bad, and this is necessary to make it less bad. And this is true of every product out there; even products, services, etc... that people hold up as "good", they still have room for improvement. If the people responsible for creating that product can't put in the time to use the product so they can feel their customers' pain and improve the product, then they should just move on.

[ItendToDisagree]

What if they have no use for the product rather than 'don't like' said product? It is quite possible to just not need to use PayPal and still work for them.