Deliver So They Come Back
Quality = performance − expectations
You delivered exactly what you promised — and the customer just shrugged. A rival delivered less and got a five-star review. The difference wasn't the work; it was what each customer expected before you started.
Satisfaction is a subtraction, not a score. Quality = Performance − Expectations. The same work feels excellent when it beats what the customer expected, and poor when it falls short of a promise you set too high. You are always judged against the number in the customer's head, never against your effort.
This is why the loudest marketing often produces the unhappiest customers. Every extra promise you make to close the sale raises the bar you must clear afterward. Win the sale by over-promising and you have quietly guaranteed a disappointment.
Selling shoes online is hard — buyers can't try them on. Zappos removed the fear with free shipping and free, no-questions returns, then quietly upgraded most orders to arrive the next day, days ahead of the promised date.
→ The unannounced early delivery became its most talked-about feature and built a reputation that helped sell the company to Amazon for about $1.2 billion.
The magic only worked because it stayed a secret; the moment a surprise becomes an advertised promise, it stops delighting and simply becomes the new expectation you must clear.
Apple's iPhone 3G was faster, cheaper, and had more features than the original — objectively a better phone. Yet many customers felt let down, because pre-launch hype set expectations so high that a few rollout glitches dropped performance below the promise. A better product scored lower on quality, purely because of the expectations attached to it.
Promise 'ready in 3 days' and deliver in 3 days = zero delight; you only met the bar. Promise 'within a week' and deliver in 3 days = a real surprise. Identical work, opposite feeling — because you lowered the expectation and beat it.
- List every promise your marketing, menu, and staff currently make.
- Quietly lower each promise to the level you can hit even on a bad day.
- Pick one small, cheap bonus you can add after the sale that nobody was told about.
- Deliver the promise reliably, then spring the bonus — and never advertise it.
The instant you print 'free next-day delivery' or 'always a free dessert' on the sign, it stops being a gift and becomes a promise you must keep forever — and a single miss now reads as a failure. Keep your best surprises off the menu.
Under-promising has a floor. Expectations still have to be high enough for the customer to buy in the first place — set them too low and nobody walks in. The goal is the highest promise you can reliably beat, not the lowest promise you can make.
Before reading on: a competitor does objectively worse work than you but gets better reviews. Name the one variable that explains it — and say which direction you'd move it.
Don't compete to raise your performance alone; manage the expectation you're measured against. Promise a little less than you can deliver, then surprise them after they've paid — satisfaction lives in that gap.
Find one promise you routinely make to customers and dial it back by a notch (a longer time window, a plainer description). Then choose one unannounced extra — a small add-on, a faster turnaround, a personal note — to spring on the next ten customers. Watch what they say.
Deliver So They Come Back