Underpromise, overdeliver: It's a common saying in business, preached by many companies (including ours), as a way to win over customers by not only meeting their expectations, but going so far beyond them they can't help but become fans.

Think of online retailer Zappos, who is famous for allowing customer service reps to spend as much time on calls as they wanted (sometimes up to six hours) or doing surprise upgrades to overnight shipping.

Underpromise, overdeliver works because it's based on a powerful, almost primal emotion: Trust. Early humans needed to work together to survive and our brains evolved to feel good during positive social interactions.

But trust isn't the only quality that makes people connect with you. The second part of the underpromise, overdeliver equation is equally as important: Surprise.

A study published by the Journal of Applied Social Psychology found how important surprise is in creating feelings of trust by asking waitstaff to give mints to customers in different ways.

Group 1: Gave mints with the check. But made no mention of mints. = 3% increase in tips

Group 2: Gave 2 mints separate from the check. Mentioned mints (i.e., "Would anyone like some mints before they leave?"). = 14% increase in tips

Group 3: Gave mints with the check. Then came back with another set of mints shortly after and let customers know. = 21% increase in tips

The third group, who not only gave more but surprised their customers, got the biggest return.

Ok, so why then, with all of this research, would I say that underpromise, overdeliver is setting you up for failure?

Simply put, the business landscape continues to change so dramatically that customer expectations--what we base our level of surprise off of--have gone through the roof.

A recent study from Social Psychological and Personality Science paid a group of puzzle doers to complete 40 puzzles, with helpers for 10 of them. The helpers, however, were told to help with either 5, 10, or 15 puzzles. In the end, the puzzle doers rated those who went above and beyond the same as those who completed 10 puzzles like they promised. Five extra puzzles was not enough overdelivery to create surprise and more trust.

The bad news? Your customers feel the same way.

When Amazon first started saying packages would arrive in 7 days, they would often deliver them in half that time. Customers loved it. But then we got used it.

You can try to reset expectations by creating your own, but existing expectations are hard to break. Doing a little better isn't enough to get attention. If you're going to underpromise, overdeliver, you need to do it at an extreme level. That's why Amazon's now trying to send you products before you even buy them.

It's nice to have the resources to overdeliver in all aspects of your business, but what if you don't have the budget? The good news is you don't always need a big budget to find optimal places to overdeliver.

We run a public cafe in our office. I've had bad customer service at many cafes in the past, like being first in line and waiting over 5 minutes before any of the staff said a word to me. So my expectations for service in a cafe are low. This was an opportunity to overdeliver without draining our resources. Our simple way to overdeliver was making sure our entire cafe team, not just the ones taking orders, knows to acknowledge people as soon as they step up to the counter.

Another underpromise, overdeliver on a budget tactic I do is my speed of follow up by email. Most people would be happy if you got back to their email in a day or two. That's the expectation of email. So I often say I'll get back within a day or two but I'm actually aiming for minutes or hours. I don't always get back in minutes or hours, but at the minimum I meet their expectation.

Underpromise, overdeliver works if you can overdeliver a lot. But if you can't reach an extreme level of overdelivery, your best bet is to at least meet expectations. Do you what you say you're going to do and research shows you'll be rewarded with a similar amount of trust and loyalty.

Published on: Jul 28, 2016
The opinions expressed here by Inc.com columnists are their own, not those of Inc.com.