Jan 22 2013
In the 1980s and 90s during the early years of the PC industry, there was alot of discussion of the economic impact of slow computers on user productivity. This was driven by some IBM research out of its Thomas Watson center in 1982: The Economic Value of Rapid Response Time by Walter Doherty and Ahrvind Thadani.
"When a computer and its users interact at a pace that ensures that neither has to wait on the other, productivity soars, the cost of the work done on the computer tumbles, employees get more satisfaction from their work, and its quality tends to improve. Few online computer systems are this well balanced; few executives are aware that such a balance is economically and technically feasible.
In fact, at one time it was thought that a relatively slow response, up to two seconds, was acceptable because the person was thinking about the next task. Research on rapid response time now indicates that this earlier theory is not borne out by the facts: productivity increases in more than direct proportion to a decrease in response time. This brief describes some of this research and the implications for increasing productivity and cutting costs that are among the chief challenges of business today."
Then ten years ago Eric Horvitz at Microsoft started looking at the impact of interruptions on PC users and how long it took them to get back on task. The entire science of interruption is interesting,. especially for the lifehacker movement that tries to deliver great productivity via various hacks and techniques to reduce interruption.
Delays and interruptions are arguably related. While you wait for a slow site to load in your browser your mind seeks something to do rather than stare at the screen, a buffering warning, or some icon of an hourglass. You switch tabs, change screens, or just give up on the poky site or service and move on to the next thing, subconsciously annoyed at your old laptop, bad internet connection, or the general crumminess of the pipe between you and your destination.
My partner Ben shared this interesting fact buried in a five-year old presentation by Amazon's Greg Linden on the economic impact of "latency" on ecommerce. This blew me away. In a presentation he claimed "Every 100ms delay costs 1% of revenue."
You can get the presentation at Strangeloop, a vendor in the site optimization space. Here are some other impacts of site speed on key performance indicators:
I'll take the liberty of extrapolating that from Amazon's most recent reported revenue of $48 billion in 2012 to mean that 100ms of performance is equivalent to half a billion dollars a year.
In my experience CTOs and CIOs at web-centric companies have tended to give more weight to availability than performance, stressing fault-tolerance and uptime over site performance. While they may be driven by how many "nines" their infrastructure is rated at, I wonder how many are making the investment in monitoring services and tools to determine their site load times. For ecommerce operations focused on converting browsers to checked out carts, I would argue response time as valid a function of success as A/B testing and strong content marketing and audience development.