In December, I wrote a blog for building awareness for web performance optimization (WPO) within digital organizations. Today I am tackling another difficult issue for these organizations: building a business case for implementing tools and processes for web performance optimization. More specifically, establishing your potential web performance ROI (return on investment). Understanding your potential ROI will help you understand the amount of resources (budget, time, and personnel) to allocate to this problem.
Establish A Baseline
The first step to establishing your web performance ROI is to identify a baseline for how your revenue-generating web pages and conversion paths are performing today. As mentioned in my previous post, there are a variety of free tools that can help you establish a performance baseline today.
The key here is identifying the most important pages to your revenue and business goals. To do this you will need the answer to the following questions:
- Which pages have the highest traffic and how much?
- What are the primary landing pages and entry points for my site?
- How are users converting? (account login portals, checkout processes, etc.)
- What are the conversion rates for key conversion funnels?
The good news is that you or someone in your organization already knows the answer to these questions. If you are having trouble finding the person with these answers, try looking for the owner of one of the following web analytics tools that capture data about your website’s end users: Google Analytics, Adobe Analytics (Omniture), Webtrends, Coremetrics, KISSmetrics, Woopra. Getting the right answers will help you narrow the scope of pages or conversion paths that should be optimized and quantify the corresponding revenue impact of website optimization.
Your next step is tying a dollar figure to performance optimizations. While every website will experience slightly different ROI numbers for performance improvements, there are a number of benchmarks and case studies published by leading eCommerce and digital media brands that can be used as a gauge for forecasting ROI numbers. Here are a few of the most notable findings:
- 2 Seconds is the threshold of acceptability for web page response times
- Google found that consumers will visit a website less often if it is slower than a close competitor by more than 250 milliseconds
- Walmart saw up to a 2% increase in conversions for every 1 second of improvement in load time.
- GQ cut load time by 80% and saw an 80% increase in traffic. Median time spent on the site also increased by 32%.
So what do all these statistics mean to you? How do you go about building a budget for performance initiatives and establishing the ROI for performance investments?
Real World Impact
Lets take a look at a real world example with data from Internet Retailer.
Home Depot’s Benchmark in 2010
According to Internet Retailer’s 2011 Top 500 guide, the average response time for the homepage of Home Depot was around 5.49 seconds throughout the course of 2010. That year, Home Depot had roughly $550,000,000 in online revenue and its conversion rate was 1.29% (conversion rate is the percentage of website visitors who buy something on the site). Here are some other metrics provided by Internet Retailer:
- Load time: 5.49 seconds
- Approx. Online Sales: $550,000,000
- Monthly Visits: 26,115,334
- Conversion Rate: 1.29%
- Average Ticket: $135
Internet Retailer also reported that Lowes.com (Home Depot’s biggest direct competitor) maintained a load time of 2.04 seconds, which is right at the 2 second threshold of acceptability for web page response times mentioned above. Most importantly for Home Depot, Lowes’ load time was 3.45 seconds faster than Home Depot’s. This is well above the 250 milliseconds Google specified as a competitive advantage.
Expected Benefits From Optimization
Based on these statistics we can recommend that Home Depot would benefit from employing optimization efforts to reduce the load time. Ideally, they would want their homepage to get within 250 millisecond of Lowe’s homepage and closer to the threshold of consumer acceptability for web page response times.
What can we expect to gain from these efforts? If we shoot to improve the homepage load time from 5.49 seconds to 2.29 seconds (right at the 250 millisecond barrier to remove Lowe’s competitive advantage) then that is a reduction of 3.2 seconds. According to Walmart’s findings (2% increase in conversions for every 1 second of improvement in load time) that would lead to the following gains:
Projected Conversion Rate Improvement = Projected Improvement in Seconds * 2% Increase in Conversions
3.2 s * .02 = 0.064 %
New Conversion Rate = Old Conversion Rate + Projected Conversion Rate Improvement
1.29% + 0.064% = 1.37256%
New Monthly Revenue = Monthly Visitors * Conversion Rate * Average Ticket Revenue
$26,116,338 * 1.37256% * $135 = $48,392,425
New Annual Revenue = Monthly Revenue * 12
$48,392,425 *12 = $580,709,102
Net New Annual Revenue = New Annual Revenue – Old Annual Revenue
$580,709,102 – $550,000,000 = $30,709,102
Note: If we use the numbers provided by Internet Retailer to calculate Home Depot’s annual revenue (26,115,334 *1.29% *135) the annual revenue produced would be $545,758,249 which would make the net new revenue from performance optimizations $34,950,853. This means that the real ROI from performance optimizations is even higher than we’re projecting based on Walmart’s findings of a %2 increase in conversions for each 1 second improvement.
Calculating Web Performance ROI
The number that we have calculated above $30,709,103 is the projected gain from produced by investing in WPO efforts and reaching a performance goal of a 2.29 second homepage load time. We can use the gain from investment to build our own budgets for implementing web performance tools and processes.
We can calculate the ROI of any technology or business investment with this formula:
ROI = (Gain from Investment – Cost of Investment) / Cost of Investment
With our above example, if Home Depot invests $1mm into WPO efforts that would lead to an ROI of approximately 97%
($30,709,102 – $1,000,000) / $30,709,102 = .9674
For the above example we only looked at the homepage for Home Depot, but to build a more thorough business case you may want to look at other key entry points and conversion paths for your website when calculating the revenue impact of WPO. This example also considers these concepts as they apply to online retailers. If we replace the “Average Ticket Revenue” with the average revenue per conversion, we can extend this model to media or other major businesses that make revenue online.
Now that you know how to build awareness of web performance culture within your business and how to calculate an ROI to make a strong business case for optimization, you’ll be ready to move forward to determine the appropriate technology tools to track and quantify performance improvements.