Why are restaurant web sites so bad?

I agree, that's why it's Flash too. Beginning designers statistically think the flashier (no pun intended) the design, the better.

Also, for a restaurant a website is really an extension of their menu/business card.

Frankly, the websites that a lot of restaurants have are perfectly fine. Even the ones made by hucksters. Is the address and phone number clearly marked? A map? Menu? Wine list? If you wanted to get fancy you could update it regularly with events and specials, but that's not really necessary.

I think you underestimate the powers that have given rise to the current state of affairs. Ecosystems of sellers usually develop around the characteristics of buyers. Restauranteurs are (as a general rule) terrible buyers and terrible clients.

You might look at a site and think "that'll take x hours to do," but you need to factor in the time that it takes to get the client there. In my experience, that's several times x. Just making the sale could be several times x.

You need to realize you are starting from scratch and that from where you stand, it's hard to know where scratch is. You will need to convince them why flash is bad (even though everyone they ask thinks it looks great). This is just one example. They will not be able to evaluate the site from the perspective of the user, that kind of thinking takes practice.

Then crisis strikes when your offer which includes some little mini-app to do that thing they want with the menus (already done) and a custom design (hired a mid range designer) costs more than the other offer which is largely auto-generated.

The reasons are similar for large companies having bad software, they are bad buyers.

I think there are a few things here that cause this:

a) You're going against the worst kind of incumbent- the 15 year old nephew or kid next door. He's cheaper, good enough, and "trusted". He is competition that exists a million times over.

b) Even if your system is better, they just can't be bothered to even get a good demo. You definitely need to build something so usable and good it's frightening, but good luck getting the full demo.

c) The guys that do have the ability to get in with a demo (GoDaddy, yellow pages,etc.) have no clue how to sell something good. What they usually sell is usually crappy, so now you have a fearful restaurant owner.

d) To crack this market it's a "heavy startup". It requires lots of capital, it requires sales, and it requires marketing. That's something guys like us don't like to often venture into. It's the same reason GroupOn requires lots of capital and other business like ReachLocal have feet on the street.

e) Churn is miserable. If you're on a monthly fee or provide additional services, good luck keeping the customer. ReachLocal has something like 80% churn,not retention, fucking churn.

Here's how I'd go about this for months 1-9. Crack the nut of an easy to use system (point b) and find distribution channels that already exist (point c) such as local newspapers, print shops, and others with a foot into the local market. You'd also need to have a system of people to crank out onboarding in a very cost efficient and systematic manner.

Hey, I'm working on something like this too :)