[Travis]

I agree. I'm also a little surprised at the false dichotomy presented in this thread. Using hand-wavy or made up numbers is not a meaningful way to have this discussion.

Every business has a different decision to make. It feels like this thread is thinking of it like maximizing an equation (FindingMaxRevenueGivenPriceSensitivies()). Instead, some products lend themselves to low cost support; sell those more cheaply. Some products require insurance-like support contracts -- those are going to be enterprisey. Some products will lend well to cheap inexpensive versions and high end enterprise versions, without changing the product (the value prop is really high to a business, esp if it's a service that can automate out the cost of a person's salary).

This thread and the obsession with A/B testing result from a hyper focus, like you said, on finding local maxima. I think it's often useful to take a step back and find the high level factors and built a pricing model around that.

(Of course, nothing excuses you from talking to customers.)