Sign In

Internet Performance Delivered right to your inbox

Case Study: Halving Page Load Times For A Major eComm Site

Slow page load times. They annoy you when you’re watching a show on Netflix, or trying to look at a friend’s vacation photos on Facebook. Now try to imagine you’re the CIO of a major eCommerce platform generating tens of billions of dollars a year, and your customers are seeing page load times of more than 600ms. Ouch. High latency and page load time is an issue faced by businesses of all kinds, but the eComm industry is without a doubt one of those to see the most immediate financial impact of high latency.

Slow page load times have revenue impacts that range from immediate (cart abandonment) to long-lasting (frustration with the brand itself), and therefore the stakeholders in fixing these issues include people like the CIO and the CMO. As soon as people on this level get involved, things move quickly when looking for the right solution.

One of the world’s largest eCommerce platforms came to Dyn seeking a solution to slow DNS responses — when you’re responsible for tens of billions of dollars in revenue annually, milliseconds count. At the time, they were seeing consistently high page load times for customers in the United States, and that was unacceptable.

Dyn was able to help this company with a secondary DNS solution and provide improved response times, leading to faster page loads in the United States. DNS latency improved by 90% to under 80ms, which in turn led to more than a 50% reduction in page load times, which has a direct impact on revenue.

Many people think having a secondary DNS provider is only in case of disaster (DDoS takes one provider out, the secondary is there to tide you over and make sure you have no downtime), but a multi-vendor approach carries other benefits to performance and reliability as well. Once you enter a secondary nameserver into the delegation, recursives check the speed it takes to resolve a query to every nameserver and gains affinity for the fastest. This controls how much traffic each nameserver will get, and ensures that traffic will be sent to the provider with the best performance at that moment.

60% of organizations last year said that downtime could cost them more than $1,000 a minute — while the giants in the industry can see downtime costing them as much as $190,000 a minute. While latency isn’t strictly downtime, if a page takes too long to load, the user may leave your site for a competitor’s before it does, which is the same effect as the page not loading at all.

High latencies cost revenue and hurt brands: don’t let yours be one of them.


Share Now

Whois: Tom Snow

Tom Snow is Analyst Relations Specialist at Dyn, an Internet Performance Management company.