Page load time and revenue - the data
September 2019
Yesterday we comforted ourselves that, despite what the interwebs say, every second of page load time does not necessarily cost us $1.6 billion.
But there’s got to be a cost, right? And there must be some reliable data about that cost, right?
For sure. Check out the Akamai State of Online Retail Performance report (PDF). They aggregate data from 10 billion user visits to explore the relationship between page speed and conversion rates.
The report is full of fascinating (and possibly disturbing) facts like this:
While almost half (47%) of all consumers browse via their phones, only 1 in 5 complete transactions on mobile.
or this:
Just a 100-millisecond delay in load time hurt conversion rates by up to 7%.
That second quote should be giving you a panic attack, so let’s dig a little deeper.
About the study
Akamai conducted the study on opted-in customers in the online retail space. (So if your site is about software signups or lead form submissions, your results may be very different.)
Akamai customers are those who have a budget for a paid CDN and managed cloud services - think big names like Walmart and Staples. (If that’s not you, your results may be very different.)
But most importantly, this study is about aggregating page speed data from across sites and sessions, putting it into buckets, and comparing conversion rates.
It’s not about the impact of speeding up or slowing down a single site.
What the study says
So what’s the deal with that 100-millisecond delay and the 7% drop in conversion rate?
What it really means is that if you take a bunch of user sessions across a bunch of sites, measure the average page load time of each session, then group them into different buckets for every 0.1 seconds, you’ll observe a trend.
Specifically, as average page load time per session gets longer, the conversion rate gets lower.
In particular, there’s a dramatic, headline-worthy, 7% drop on mobile when you move from the 2.7 second bucket to the 2.8 second bucket.
What it doesn’t say
This study does not mean that reducing your page speed by 0.1 seconds will increase conversion rates by 7%. It doesn’t even say that you’ll see a drop in conversions if your site slows down.
It’s sure phrased that way, but that’s not what it says. It simply tells us that on average, among CDN-using online retailers, user sessions with faster page load times tend to have higher conversion rates.
We’ve got to watch out for the third cause fallacy here - sessions and sites with faster page speed have higher conversion rates, sure, but does the speed cause conversions?
Or is it possible that investing in page speed is correlated with investing in conversion optimization and UX, meaning that the speed and conversions are both caused by “having money”?
What to do
So far all we can definitively say is that overall, faster sites tend to have higher conversion rates. And intuitively, we know that humans tend to be impatient. And we know that we hate it when a page takes 20 seconds to load.
But should you invest in improving your page speed? How much ROI can you expect from doing so?
The jury’s still out. I recommend you do nothing for now 😁.
Let’s look at some anecdotal evidence, and try to arrive at a strategy that acknowledges the truth in the trend, while ignoring the hype in the headlines. See you tomorrow! 👋