4 Tips to Help Your Website Make Money This Holiday Season
Many retail websites during the 2010 holidays failed to meet the performance demands of consumers who expect web pages to download in two seconds or less — even during peak traffic periods. A recent analysis of web page visits found that end user abandonment rates rise 8 percent with an extra two seconds of wait time, meaning that slow websites can have a negative impact on revenues and profitability, as well as brand and customer loyalty.
This year’s underperforming retailers share three common mistakes:
- load testing isn't properly prioritized;
- key stakeholders aren't involved in the load testing process; and
- load testing isn't conducted from the end user’s perspective.
In contrast, this year's most successful retailers have taught us several key lessons (applicable to any peak traffic event) which can help others realize a profitable holiday season. These lessons include the following:
1. Early and thorough planning is critical to your online success. You must allocate the appropriate time and effort to execute an effective load test. Otherwise, your load testing is rushed, ineffective or incomplete. These tests must take into consideration what pages, applications, customer segments and geographies need to deliver the best performance. By solving these problems first, you'll engineer a more streamlined, strategic load test that identifies the most critical (and potentially costly) bottlenecks. You'll also have adequate time to address these issues.
2. Both marketing and IT execs must be considered in the plan. Before executing a load test, define what constitutes acceptable performance. Conduct the test based on those parameters. Representatives from your marketing team should be involved so they can share information on anticipated sales and growth projections in order for the load test to be scaled properly. Likewise, representatives from your IT department must communicate the capabilities and limitations of the environment, not so much in terms like network capacity and CPUs, but how heavy loads will impact the end user experience given the realities of the current infrastructure.