“Eating your own dog food” means using your product everyday, as your clients will do. I was thinking about this today and wondering how this rule applies for different products.

I concluded there are three categories of products, as far as dog-food eating is concerned:

A) Products targeting consumers or developers / software companies (you belong to the target group from the start)
B) Products targeting existing businesses. This is the case with CRM, Helpdesk and other products requiring an existing business (you belong to the target group after a point in time).
C) Products targeting a niche that has nothing to do with you or your company. For example think of products for real estate agents, insurance salesmen or Internet / Cyber cafe management.

So while category A is easy, categories B and C are more problematic. Initially category B is the same as C, but merges with A after you get your v1.0 out and acquire some clients. On the other hand, C will always be a problem.

How can you continuously improve a product that you don’t use everyday? The only answers I can think of are:
1) have strong domain knowledge
2) frequently get feedback from interested parties
3) get someone else to work closely with you on this. You must trust his judgement and make sure he explains all reasons for asking a change or a new feature, so that you too, can understand the reasons.

Of course this is not exactly eating your own dog food. It feels more like testing, but then again I can’t think of other options.

Personally I believe I have enough knowledge to come up with v1.0 on my own. But this is the wrong mentality and the whole reason for eating your dog food ;-) So, I intend to also go with both option 2 and 3.