Client Case Study: How speed equals conversion

I’ve written plenty of blogs about the importance of speed to a website and the website’s commercial performance and prospects.

The faster a website performs – quite literally the speed with which it operates – the more the users that will make it through the website; what is measured as and referred to as the website’s conversion (or acquisition) rate.

There are plenty of other factors at play when it comes to why or why not a website is converting, though speed is a great one to focus on for a few reasons:

1. Improvements in speed can be immediately and clearly measured.

2. There is no theories about the impact speed has on conversion: the faster the website, the more the sales (conversions) you’ll get. Immediately.

3. Often, there are quick and inexpensive wins: reducing page weight (by reducing image sizes and removing unnecessary content), removing slow third-party tracking code, optimising a truly offensive database call, or whatever the quick wins are.

4. It can possibly force possibly an uncomfortable conversation with the hosting provider, though that is a good thing: it opens eyes to the benefits of quality hosting (and the false economies of poor hosting) and might lead to a longer-term and worthwhile migration. (Get some advice on this first: a lot of people race of the jugular of their hosting company at the first mention of speed whereas hosting and infrastructure may not at all be the issue).


A good web developer should be able to do some performance profiling of your website quickly and inexpensively and working out a plan to improve your website’s performance should be straightforward: work out the quickest wins, work out time to execute and go forth.

And repeat and repeat the process until the wins are exhausted and the law of diminishing returns kicks in.


What numbers am I looking for?

What are good numbers, what are bad numbers and what sort of return should you see?

Google’s golden rule is that any page taking over two seconds to load, needs work. There is plenty of evidence that much more than that and you will see big drop-offs in users.

Certainly, nobody is going to stick around for 10 seconds waiting for a web page and certainly not a whole website as they trudge around trying to make a purchase with you.

Lot's of stats out there to support it; indeed, one article I read this morning comfortably (and probably accurately) quoted that every 1 second page load delay saw a drop of 7% in overall website3 conversion: 11% lower page views and a drop of just over 15% in user satisfaction.

(Not to speak of the impact having a slow website has on SEO and your rankings in Google...)

All these points were a catalyst for this blog: an incident with a large retail client this morning. Good website, good traffic, good range of product.

Due to a requirements of their analytics platform, we store enormous amounts of previous session data and over the past 12 months it has really built up, to the extent that we were seeing database performance degradation issues.

In other words, the website started slowing down and wouldn’t you know it, so did the website’s overall rate of conversion.

A few hours of fixes and the website is back to its former glory.

Overall website conversion jumped immediately and was up a number of percentage points. Every second of page improvement saw a correlating improvement in website sales. Every one second we clawed back lead to a few percentage points in increased conversion. At least.


We could spend our lives improving the usability of our websites, improving the content, playing with product price points and all sort of manner of clever tricks for wringing out conversion and users. And we should be doing all of this and then some.

Just don’t forget speed and the instant impact it can have when achieved. It will work.