What is “Dogfooding”?

So there’s an important concept in the tech world that has a very strange name… I tell clients about the concept frequently, and usually get laughter when I mention the term: dogfooding.

Say what? What is dogfooding, anyway? Yes, it’s an awfully strange term, but it makes perfect sense. Here’s the idea: if you work at a dog food company, and you have a dog, at the end of the day, you’d better be bringing home your company’s dog food to feed your dog. Why? Because if your product is as good as your marketing says it is… you would buy it yourself. [Note: I realize this analogy breaks down as some point, as all analogies do. Most Gulfstream Jet mechanics can’t afford the luxury private jets they work on all day. But in this particular case, with dog food, it fits perfectly.]

Below, I’ll give you three examples of the concept: a good example, a bad one, and a really strange one.

Three Examples of Dogfooding in Action

…

Good Example

The best example of dogfooding I know of is by 37signals. They’re a software company that makes business productivity software, and they use their own software to run their software business. As it should be. It’s brilliant marketing. It shows confidence in the products they make. It also makes them much better at sales, support, and quality control. Because they use their products every day, they know when it doesn’t work. They know immediately if a software update breaks something. They know how annoying it is that there isn’t a “such and such” button where they’d expect it. So when a customer calls in and expresses frustration about a particular issue, the support tech feels the frustration. And they can better convey to the programmers what needs to be fixed, and why.

Bad Example

A perfect example of a lack of dogfooding is at car dealerships. A few years ago, my car completely died and I needed another one. So, not knowing where to start, I went to a car dealership. (This turned out to be a major mistake, for several reasons, but I’ll save that for another post). You know what I noticed as I browsed through the rows of cars? Each dealership has a “staff only” parking lot, where the employees park when they get into work in the morning. And you know what? At the Ford dealership, I saw Acuras in the employee parking lot. At the Volkswagen dealership, I saw Toyotas. Why? Because even though a salesman will hawk the merits of owning, say, a GMC Sierra all week long, at the end of the day he’ll clock out, and drive home. …in a Honda Civic. Of course he doesn’t drive a GMC. Because he doesn’t want a GMC. Even if he does, at a minimum, he doesn’t want it enough to actually buy one. He’ll tell prospective customers that it’s worth the investment, but he’s not buying his own pitch. He’s not investing his own money.

Isn’t that fascinating? I think the dealerships should be embarrassed. When I was out searching for a car that day, what I should have seen was Chevrolet’s employee lot packed full of brand new Cavaliers and Trailblazers. But I didn’t.

Strange Example

The strangest example of an attempt at dogfooding I know of is at Microsoft. In their Headquarters in Redmond, Microsoft (allegedly) put out a big acrylic bin for their employees to “turn in” their iPods and all their treasonous sins would be forgiven. Why? Because Microsoft was building a product—with a cringeworthy name: the Zune—that competed with the iPod. The Zune is one of the biggest personal music player flops ever. There are some people who will defend the Zune as being a good media player, and even (correctly) reminding us that it has features the iPod has never had, such as an FM Radio and wireless syncing. But none of that matters. What matters is that the massive technology behemoth from Washington that spawned the once-richest man on earth, sold a grand total of 2 million units. The product they were trying to compete with—Apple’s iPod—has sold 297,000,000 units. If you’re into big numbers, that means the world has bought over 148 times more iPods than Zunes.

How is this an example of dogfooding? It shows you how not to do it. Microsoft had to try to convince their own people to try their own product. Or at least to not try the competitor’s. Even though I’m sure the clear bin for contraband Apple products was a joke, the joke is on Microsoft. The real world told us what we need to know: nobody wanted a Zune.

My Point

I think eventually, everyone in the sales or marketing industry has to pitch a product or service they don’t necessarily believe in. Perhaps even many times. And that’s ok. It’s not the marketer’s fault. It’s the manufacturer’s. If their product was so fantastic, their employees would want it. And if they really believed in their product, they would remove any barriers that keep their employees from owning and using the product. I think 37signals’ approach is right on—the person best qualified to give you feedback on your product is he who owns it. And if he who owns it is also he who created it, the product will improve automatically.

Ideally, a BMW dealership would offer their sales staff a used BMW as a company car. Microsoft and other offbrand-music-player-makers would simply give their employees the product they’re selling. It’s good business. It may cost them a few hundred dollars per employee, but that’s cheap. What’s expensive is conceiving an idea for a product, pitching that idea to investors, building a factory, hiring workers, designing, manufacturing, and then marketing and selling a product that nobody wants.

If you have a dog, and you work at a dog food factory, please bring him a doggy bag each night. Either literally or figuratively. And if your dog won’t eat what your company makes, please seriously consider quitting your job and go work somewhere else. It’s a matter of integrity. And your dog’s health.