We know that slow page loads drive people away. In recent years users expect web pages to load in 3 seconds or less - even on mobile devices. Speed is considered the most important feature by some. So, when I hear stories about front end performance not being a business priority when it accounts for about 87% of the time a page takes to load I see we have a need to help our business decision makes understand the importance. Let's take a look at some information and resources that might help.
Business decision often care about features, ads, or some other functionality in the site. How it works or how fast it's delivered is often a second though if one at all. It's not always easy to understand since it's not something you can click on or stare at how pretty it is. That by no means makes it non-quantifiable. Let's try to quantify front end performance in a way a business decision maker can understand.
Assumption: Where Performance Is Experienced
First, I want to address an assumption which is more observation than assumption. Front end performance is the majority of the performance time experienced by an end user. This has been observed and measured by the HTTP Archive and Strangeloop that observed 97% of the time is in the front end for mobile devices.
Business decision makers, in my experience, like bullet pointed reasons with references. They also like measured numbers that can point to more money or engagement.
Increased User Interaction
- Bing found an increase in page load time of 2 seconds caused about a 2% drop in queries per user. The time to first click was about double the time of the delay. An additional delay of 500ms caused over a second of additional delay before the first click.
- Shopzilla saw a pageview increase of 25% by dropping 5 seconds from pageload time.
- Mozilla shaved 2.2 seconds from their pageloads and it increased download conversions by 15.4%.
- Yahoo found a 400ms slow down caused a 5-9% drop in traffic.
Financial Reasons - Show Me The Money
- Bing found an increase in page load time of 1 second caused a 2.8% loss in revenue per user and a 2 second delay cause a 4.4% loss in revenue per user. Even a .5 second increased delay caused a 1.2% loss per user.
- Shopzilla saw a conversion rate increase of 7-12% by increasing performance by 5 seconds.
- Amazon found "Every 100ms delay costs 1% of sales". - This was one of the earliest reports on performance.
While you may need more data (and it does exist), hopefully information like this will help spark an interest from decision makers.
These all have references you can follow. If you want even more reasons (though without references) checkout the post from Blaze.io on this same topic.