There is nothing more frustrating than a program that generates an error, asks you to try again, and continues to show that same error message, trapping you in an endless and frustrating loop. Even for experienced users, it can be hard to figure out when to try again and when to just give up. But these decisions don’t have to be so difficult.

Here’s a case in point. When I tried to e-file my state tax return, it failed with an error message that listed a bunch of possible causes for the error. I took a shot at correcting the issue, waited a few hours for the e-file response, and got the same error. Third time’s the charm, right? Nope — failed again. As I learned later, I needed to mail in the return due to obscure and stupid issues with the state’s e-file system.

While I was looking into the error message, I noticed that many other people had reported nearly identical problems. And virtually every single one of those people ended up having to mail in the return. Then it hit me: if the software provider has that data, why not use it? In other words, if you have data on every customer that received that error, how many times they retried, and whether they were eventually successful, then use it!

Instead of blindly asking the user to try again, you could use peer data to show a message like this: “78% of customers who received this error were able to resolve the problem by trying again. Would you like to review your information and try again?” In my case, it would have showed that something like 95% of customers never got past the cryptic error message, indicating that I shouldn’t have wasted my own time. By making information on peer success rates available to other users, you can save people time and frustration — and give them a great reason to tell others how valuable and insightful your product is.