Previous research has shown that user frustration increases when page load times exceed eight to 10 seconds, without feedback (Bouch, Kuchinsky, and Bhatti 2000, King 2003)., Newer evidence shows that broadband users are less tolerant of web page delays than narrowband users. A JupiterResearch survey found that 33% of broadband shoppers are unwilling to wait more than four seconds for a web page to load, whereas 43% of narrowband users will not wait more than six seconds (Akamai 2006).
Tolerable Wait Times
In a 2004 study, Fiona Nah found that the tolerable wait time (TWT) on non-working links without feedback peaked at between 5 to 8 seconds (Nah 2004). Adding feedback, like a progress bar, pushed the TWT to an average of 38 seconds. Subsequent attempts at non-working links revealed lower TWTs, peaking at 2 to 3 seconds without feedback. Nah concluded that the TWT of web users peaks at about 2 seconds. With regard to behavioral intentions to return to a site, Dennis Galletta and others found that they level out at 4 or more seconds and attitudes flatten at 8 or more seconds (Galletta et al. 2004).
The Effects of Slow Download Times
Even small changes in response times can have significant effects. Google found that moving from a 10-result page loading in 0.4 seconds to a 30-result page loading in 0.9 seconds decreased traffic and ad revenues by 20% (Linden 2006). When the home page of Google Maps was reduced from 100KB to 70-80KB, traffic went up 10% in the first week, and an additional 25% in the following three weeks (Farber 2006). Tests at Amazon revealed similar results: every 100 ms increase in load time of Amazon.com decreased sales by 1% (Kohavi and Longbotham 2007). Experiments at Microsoft on Live Search showed that when search results pages were slowed by 1 second: (Kohavi 2007)
- Queries per user declined by 1.0%, and
- Ad clicks per user declined by 1.5%
After slowing the search results page by 2 seconds:
- Queries per user declined by 2.5%, and
- Ad clicks per user declined by 4.4%
The Effects of Slow Response Times on User Psychology
Slow web pages lower perceived credibility (Fogg et al. 2001) and quality (Bouch, Kuchinsky, and Bhatti 2000). Keep your page load times below tolerable attention thresholds, and users will experience less frustration (Ceaparu et al. 2004), lower blood pressure (Scheirer et al. 2002), deeper flow states (Novak, Hoffman, and Yung 2000), higher conversion rates (Akamai 2007), and lower bailout rates (Nielsen 2000). Faster websites are actually perceived to be more interesting (Ramsay, Barbesi, and Preece 1998) and attractive (Skadberg and Kimmel 2004).
The old 8 to 10 second rule has diverged into the haves and have nots. Broadband users expect faster response times, while narrowband users have been left behind. As broadband becomes more widespread the size of the average web page has increased to over 300KB and the average number of objects has increased to over 50 per page. Users experience psychological and physiological effects when interacting with web pages, experiencing frustration when not completing tasks and engagement at faster web sites. Narrowband users are suffering the most from the speed tax of objects that now dominates most web page delays. Increasing the speed of your site will improve your conversion rates, reduce shopping cart bailout rates, and make your site more appealing to users.
- Akamai, “Boosting Online Commerce Profitability with Akamai,”
- Akamai Technologies, 2007, http://www.akamai.com (May 30, 2008). Based on the finding that 30% to 50% of transactions above the 4-second threshold bail out, Akamai estimated that by reducing the percentage of transactions above this threshold from 40% to 10%, conversion rates will improve by 9 to 15%.
- Akamai. June 2006. “Retail Web Site Performance: Consumer Reaction to a Poor Online Shopping Experience.”
- Akamai Technologies, (accessed May 30, 2008). This is a JupiterResearch abandonment survey commissioned by Akamai.
- Bouch, A., Kuchinsky, A., and N. Bhatti, “Quality is in the Eye of the Beholder: Meeting Users’ Requirements for Internet Quality of Service,”
- in CHI 2000 (The Hague, The Netherlands: April 1-6, 2000), 297-304. Found that latency quality ratings drop off at around eight to 10 seconds.
- Ceaparu, I., Lazar, J., Bessiere, K., Robinson, J., and B. Shneiderman, “Determining Causes and Severity of End-User Frustration,”
- International Journal of Human-Computer Interaction 17, no. 3 (2004): 333-356. Slow websites inhibit users from reaching their goals, causing frustration.
- Farber, D., “Google’s Marissa Mayer: Speed Wins,”
- CNET Between the Lines, Nov. 9, 2006, http://blogs.zdnet.com/BTL/?p=3925 (May 30, 2008).
- Fogg, B. J., et al., “What Makes Web Sites Credible? A Report on a Large Quantitative Study,”
- in CHI 2001 (Seattle, WA: March 31-April 5, 2001), 61-68.
- Galletta, D., Henry, R., McCoy, S., and P. Polak, “Web Site Delays: How Tolerant are Users?”
- Journal of the Association for Information Systems 5, no. 1 (2004): 1-28.
- King, A., 2003, Speed Up Your Site: Web Site Optimization
- Indianapolis: New Riders, 2003, 25. Found an average of 8.6 seconds for tolerable wait time.
- Kohavi, R., “Practical Guide to Controlled Experiments on the Web: Listen to Your Customers not to the HiPPO,”
- (Washington, DC: eMetrics Marketing Optimization Summit, Oct. 20-24, 2007), 12, http://exp-platform.com/Documents/2007-10EmetricsExperimenation.pdf (Feb. 13, 2008). This talk contains Microsoft Live results.
- Kohavi, R., and R. Longbotham, “Online Experiments: Lessons Learned,”
- Computer 40, no. 9 (2007): 103-105. The Amazon statistic was taken from a presentation by Greg Linden at Stanford: http://home.blarg.net/~glinden/StanfordDataMining.2006-11-29.ppt.
- Linden, G., “Marissa Mayer at Web 2.0,”
- Geeking with Greg, Nov. 6, 2006, http://glinden.blogspot.com/2006/11/marissa-mayer-at-web-20.html (May 30, 2008).
- Nah, F., “A study on tolerable waiting time: how long are Web users willing to wait?”
- Behaviour & Information Technology 23, no. 3 (2004): 153-163.
- Nielsen, J., Designing Web Usability
- (Indianapolis: New Riders, 2000), 48-49.
- Novak, T., Hoffman, D., and Y.-F. Yung, “Measuring the Customer Experience in Online Environments: A Structural Modeling Approach,”
- Marketing Science 19, no. 1 (2000): 22-42.
- Ramsay, J., Barbesi, A., and J. Preece, “A psychological investigation of long retrieval times on the World Wide Web,”
- Interacting with Computers 10, no. 1 (1998): 77-86.
- Scheirer, J., Fernandez, R., Klein, J., and R. Picard, “Frustrating the user on purpose: a step toward building an affective computer,”
- Interacting with Computers 14, no. 2 (2002): 93-118.
- Skadberg, Y., and J. Kimmel, “Visitors’ flow experience while browsing a Web site: its measurement, contributing factors and consequences,”
- Computers in Human Behavior 20, no. 3 (2004): 403-422.