Is it true that price comparison websites, particularly flight comparison websites 'delay' the search results of price information so users feel that the search is comprehensive?
Otherwise what is the explanation for a 5 or 6 second delay for the results - surely looking up that information does not take that long?
Therefore is this a user friendly experience, so they feel like the search has been comprehensive, even if search results delivered instantly are the same?
The delay is due to the fact that since the flight information must be up to date both price-wise and availability-wise and not based on a cached database, as web search engines do, the server must contact multiple external services - those of all the flight companies to get up to date info, thus the delay. The chances that all of the results for your specific search are cached, especially since the cache must be short living (a day old flight availability is too old, whereas a day old webpage is reasonable), not to mention that technically there is a storage limit for the cached results which is probably much smaller than all the possible results (assuming the search company doesn't take high fees for covering a huge cache storage).
I am sure they appreciate the fact that some people think they are much faster than they really are and that the delays are actually only for sake of a dramatic effect :-)
On one system I worked on, a predecessor had added a 500ms delay to an AJAX response. I asked why and learned it was because the results came back so fast the screen often just jumped and it felt strange. I added that to my bag of tricks and think it's a reasonable application of a forced delay.
As a student of Dan Ariely, I learned that thinking rationally about UI or anything else is not always best practice. It would not surprise me if people did prefer a little lag... particularly in the airline case since the result is a "payoff". A little anticipation may add to the experience.
Flight comparison websites typically query the merchant websites through APIs, but it also may occur that they need to "manually" crawl the hit page of a flight company and extract the flight details with prices. Crawling includes some extra round into the search, and these easily add up.
Another aspect to consider, that some flight comparison websites have a tendency to obfuscate the hits from low-cost carriers (at least in Europe), and hide them among the searches. This can be done more subtly when the search results appear continuously.
No, no, no.
I currently work with an engine for searching flights.
Basically you got two major variations:
You have to understand what underlying technology is used. If the flight comparison uses other sites (it's a meta search site) it will cache all entries and thus they use a cache mechanism. But you need to understand that such a site can have around 10 000 new cache entries per second so even if that meta engine searches in cached entries there's still somewhat of a huge overhead per search.
You also have to understand that a simple search can result in a couple of million possible alternatives. This is true for both the conventional real-time search made through a GDS but also for searches against a cached database.
In short: Yes, some add delay but the most delay you actually see is due to the amount of data that is actually being processed.
