It sounds like you’re describing what some companies call “delight” and there’s been a book or two written on the subject. At the core of both “surprise” (which could be positive or negative, a great enhancement, btw) and “delight” (which would be positively valenced) is the notion that there is some expectation set and an event exceeds (or falls below) that expectation. A principle in sales suggests that you underpromise but overdeliver — presumably to take advantage of exceeding an expectation.

The challenge for this effort is usually understanding where customer expectations are — and population heterogeneity makes this excruciatingly difficult. Also, customer expectations are both situational (flowers sent on Valentine’s Day are so much more expected than any other day of the year) and a moving target. Your last experience waiting for the rep to answer your call at your bank certainly sets the stage for how well that customer reacts to YOUR customer experience.

I like how you related this thought to the idea of committment. There’s a larger body of work relating committment to preference and also the “stickiness” of the relationship. Imagine a two-by-two matrix. Along the vertical axis is whether you prefer a brand above others (high) or not (low). Along the horizontal axis is the pain you would experience from the loss of use of that brand (far right is high, far left is low). High preference and high pain puts you in the “Committed” quadrant. Conversely, low preference and low pain puts you in the Exiting quadrant. Moving customers around from the other two quadrants, Trapped (low preference, high pain) and Switcable (high preference, low pain) to the Committed quadrant requires either making their stickiness higher or having them prefer you more.

What better way to do this than to “delight” or surprise your customer?