Seth Godin wrote about the Pump in New York last week. It's a 350-square foot lunch spot that offers "precisely 41,000,000 combinations" from 50 ingredients. There's a line out the door everyday; Seth concludes these are people who love customization, people who wouldn't wait around to get into restaurants with "no substitutions" policies.
That makes sense. During my last year at EV1Servers, I never managed to clear out the custom order inquiries queue. Now that I think about it, we probably offered more hardware/software combinations than the Pump's lunch options. I sold servers while waiting in planes for takeoff, standing in grocery store checkout lines, petting my neighbor's dog... All it took was picking up the phone.
Just now, though, I came across Eric Allam's post on limited choices. Eric has just launched FizboFlorida, a real estate listings site with 3 service plans. He cites Freshbook's Michael McDerment on why small numbers of buckets work better than infinite customization:
What is better: buckets or custom pricing? Buckets. How do I know we learned this? Since changing the pricing page on our site, our sign-ups/trails have increased 30%... What’s amazing is our actual prices are identical, but just by presenting our pricing in three easy to understand buckets, conversions of first time visitors to trials have increased about 30%... Basically we had been trying to please everybody and therefore we were confusing people with our pricing plan.
That makes sense too. My friend Irakli was just telling me that he hates ordering from too-long wine lists. Just offer me something, he said. Don't make me think through endless possibilities. Having managed a high end audio store, he's concluded that sellers - if they want to stay in business - are responsible for helping buyers narrow their options.
So what's really better? Both, probably. EV1 has a list of pre-configured servers on its website, and I'm sure the Pump has a set of recommended combinations. For every customer who insists on picking his own ingredients, there might be plenty more who'd gladly order from the menu. The question is, what's the incremental cost of pleasing Seth's "yes substitutions" set?