Ernst Abbe was a German scientist, known for his work on microscopes at Zeiss.
He gave his name to Abbe errors - the way angular errors of measurement are magnified by distance. You may have noticed when you play snooker that a tiny error in the angle at which you hit a ball results in missing the pocket by a huge margin, and much the same thing happens when you measure errors.
It's because of these errors that the Universal Principles of Design includes the Abbe principle:
"Measure things as close to their action as possible."
Measure close to the event
It's a rule that works just as well for customer insight. Most customer experience researchers will tell you that you should run event-driven surveys as soon as possible after the event we're asking customers to talk about. They may add rules of thumb, which (somewhat suspiciously) have tended to shrink as improvements to survey technology make it possible to get closer and closer to the event. Where we used to talk in terms of 1-2 weeks, we'd now talk about days or even hours. The closer the better, with the proviso that it's important to make sure that the customer feels that the event is over before asking them how well it went.
What's close enough?
So what does the science of memory actually teach us about the way memory works? In their landmark book "The Psychology of Survey Response" Tourangeau, Rips, and Rasinski explore the topic of recall in surveys in great detail. As you'd expect, time is a big factor...but it's more complicated than that. You can probably remember the names of your classmates at primary school better than what you had for lunch last Tuesday. So what else do we know?
In other words, the more everyday an occurrence, the more important it is to survey customers as soon after the event as possible. Reminding customers of some of the details of what happened will help to make their memories about how they felt more accurate.
For common, low-impact, events: survey soon and provide cues to help accurate recall.
For important, emotionally-involving, events it's less important to survey customers instantly, and the need to make sure that the event has properly finished in the customer's mind takes precedence. For example, if you want to understand the mortgage application journey, you can comfortably wait a week or two without worrying too much about the passage of time.
For significant events: survey only when you're sure the event is complete.