Blog · 7 November 2026 · 8 min read

What a slow website costs you: the numbers to know before you pay for optimisation

“The site works, customers call – why should I pay to make it faster?” That is a fair question, and it deserves an answer in money, not in PageSpeed points. Below are four studies that actually measured the effect of speed on sales, four ways a slow site leaks money, and a simple calculation you can do for your own business in five minutes. No promises of “+300% conversions”.

What has actually been measured

The web is full of speed statistics, most of them copied third-hand. The four below come from studies you can open yourself (links at the end). Each has limitations, and we state them.

The common thread: none of these studies talks about “doubling your sales”. They talk about a few to a dozen percent at each step – but those percentages apply to every visit, every day, with no extra advertising budget.

Four doors the money leaves through

1. Bounces before the page even appears

The simplest loss to understand and the hardest to see in your statistics: a person who closed the tab after two seconds of white screen often never shows up in Google Analytics, because the analytics script had not loaded yet. Your reports show the traffic that survived. If you paid for that click in an ad, you paid for nothing.

2. Abandoned forms and carts

The second loss is people who stayed but gave up: the “Send” button reacts after a second, the dropdown stutters, nothing happens after “Add to cart” so they click again. This is what the INP metric in Core Web Vitals measures. The Vodafone test shows the scale: same offer, same price, 11% more carts simply because the page responded faster.

3. Your position in Google

Google “highly recommends” achieving good Core Web Vitals (LCP within 2.5 s, INP under 200 ms, CLS under 0.1) and says plainly that this aligns with what its ranking systems seek to reward. It is one signal among many, not the main one: content and links weigh more. In practice it works as a tie-breaker – between two similar local-business sites, the faster one has the edge. Do not expect to jump from 30th to 3rd on speed alone; expect not to lose to a competitor over something you can fix in a week.

4. More expensive ads

Google Ads rates “landing page experience” as part of the Quality Score, and load time is one of the factors it considers. A lower Quality Score means a higher bid for the same position. Add door number 1: some of the clicks you pay for end on a white screen. If you spend a few thousand a month on ads, a slow landing page is a real cost, not a theory.

Work it out for your own site

You do not need studies – you need three numbers of your own. Take them from Google Analytics or your shop's dashboard:

1. Visits per month (sessions), ideally from phones only.
2. Conversion rate: what share of those visits ends in an enquiry, a call, a booking or a purchase (in percent).
3. Value of one conversion: average order value, or average job value multiplied by the share of enquiries you turn into customers.

An example for a service business (illustrative numbers): 3,000 mobile visits a month, 2% leave an enquiry, that is 60 enquiries; one in five becomes a job worth $800. That is 12 jobs and $9,600 a month from the website. If the site loads in 5 s today and in 2 s after the fixes, the studies above suggest a conversion lift of tens of percent. Assume a cautious +20%: two or three more jobs, so $1,600–2,400 a month, year after year. For a shop with 20,000 visits and a $45 basket the same arithmetic gives five-figure sums per year.

Now the other side of the equation. Typical fixes on a small-business site are a few hours of a developer's time (images, plugins, caching), sometimes a hosting change for a modest yearly fee, rarely a rebuild of the theme. If the yearly gain in the first column is several times the cost in the second – and with any meaningful traffic it usually is – the decision makes itself. If it is not, read on.

When speed is not your problem

• The site gets 100 visits a month. Even doubling conversion adds one enquiry. Visibility and content first, speed later.
• The site has no reason to be fast. No form, the phone number as an image, an offer from three years ago – faster loading will not fix a visitor who does not know what to do next.
• The problem is elsewhere. A fast site whose contact form sends nowhere, or whose certificate is about to expire – more common than you would think; our launch checklist lists 24 such things.
• Your PageSpeed score is 85, not 25. The difference between 85 and 100 is invisible to customers. Do not pay to chase a perfect score.

How to measure so you do not pay for a promise

1. Before you commission anything, record a PageSpeed Insights result for the home page and your main offer page – the “Mobile” tab. If you see the “Discover what your real users are experiencing” section, screenshot it: that is data from your customers' phones over the last 28 days, and it is what counts for Google.
2. Give the developer a list, not the word “faster”. The list should name specific files and plugins – where to get it is described in our post on slow mobile sites.
3. After the fixes, repeat the lab measurement right away and check the real-user data a month later – that is how long Google's measurement window is.
4. After a quarter, compare mobile conversion before and after. That is the only measurement you really care about.

Sources

• Google / Deloitte Digital, Milliseconds Make Millions (2020)
• web.dev, Why does speed matter? (BBC figure)
• web.dev, Vodafone: A 31% improvement in LCP increased sales by 8% (2021)
• Portent, Site Speed is (Still) Impacting Your Conversion Rate (2022)
• Google Search Central, Understanding Core Web Vitals and Google search results