I don't disagree with Seth Godin very often. In fact he's probably the one person (besides my wife and kids) that change my current thinking about a topic. But I disagree that there's a problem with being perfect.
His hypothesis is that being perfect is a) not remarkable (to still one of his older phrases) and b) only remarkable when you screw up.
I agree with b, as I think most people will. On a, I do believe perfection *is* remarkable. "FedEx" it is a statement of being perfect. They are the de facto standard when needing to overnight something. That FedEx is a verb in our language is a testament to "perfection being remarkable."
My favorite restaurant is Jak's in Issaquah. The best steakhouse I have ever been to, and that includes the great steakhouses of Chicago. They are simply perfect. Been there about 10 times over the last two years, and always a perfect experience. Wine is always at a perfect temperature. My steak is always a perfect cut cooked to a perfect medium rare and delivered to me at a perfect temperature. In fact my first bite of a Jak's steak is always better than I anticipate. The UFO potatoes (their specialty) are always perfect. As is the timing for the service staff, their service, and their atmosphere. It's perfect; as such we remark about it. I've even blogged about it multiple times.
However, I don't think perfection is an end-all-be-all. Perfection is a business choice companies need to make. I don't always opt for a perfect meal; in fact, I rarely do. I spend way more money on imperfect.
Here's my take on being perfect:
1. If you are going to be perfect, you have to be perfect, period. FedEx cannot screw up.
2. It's important for Product Managers to recognize when something is "good enough" and when you're overserving your market. Google Spreadsheets has adoption because it is "good enough" and Microsoft Office has overserved the market.
3. Make a strategic decision whether to be perfect. Sometimes (most times?) the goal is not to be perfect but to move the needle.
4. Understand that most customers in most situations do not value perfect.
I remember proposing an impressive appliance diagnostic solution for Whirlpool and talking to their customer support organization. The solution include communications with an appliance over the phone, Artificial Intelligence, and a voice (over phone) menu to collect data. The most challenging part was the voice interaction, and we hadn't worked out all the kinks by the time we talked to the CS team. I thought we were at a disadvantage because our voice component worked only half the time. Interestingly enough, they thought it was great because it worked half the time. Why the discrepancy? I was comparing our 50% solution to my personal goal of a 100% solution. They were looking at my 50% solution as compared to their current 0% solution. My solution moved the needle 50% for them! As such they loved it.
Clearly a case where perfection should not be a goal. And there are numerous ones in the business world where understanding what is "good enough" is the difference between success and failure.