- Google and SOASTA (2017) found bounce probability rises 32% as load time goes from 1 to 3 seconds, and 123% from 1 to 10 seconds.
- Deloitte and Google (2020) measured a 0.1 second speed improvement lifting retail conversions by 8.4%.
- You can convert a speed gain into money: revenue x conversion uplift per second saved equals the annual figure.
- Mobile visitors abandon sooner, and Google ranks pages partly on Largest Contentful Paint (good under 2.5 seconds).
- Server response time sets the floor: no front-end tuning can undo a slow, shared backend.
How does page speed affect conversions?
Website speed affects conversions directly: slower pages lose visitors before they act, and the effect is measured, not folklore. Google and SOASTA found that moving from a 1 second to a 3 second load raises bounce probability by 32%. Deloitte and Google measured a 0.1 second gain lifting retail conversions by 8.4%. Speed is a revenue lever.
The mechanism is simple. Every extra second is a fresh chance to abandon. A visitor who leaves before your first content paints never sees the product, reads the offer, or reaches the checkout. Slowness also erodes trust: people read a sluggish page as a sign of a shaky business, and they behave accordingly. The effect compounds through your funnel, where each slow step sheds a share of the people who reached it.
Deloitte and Google, Milliseconds Make Millions, 2020: a 0.1 second mobile speed improvement raised retail conversions by 8.4%.
What does a one-second delay actually cost in revenue?
A one second delay costs real money, and the earliest hard figures came from the biggest sites. Amazon reported that every 100 milliseconds of added latency measurably cost sales. Google found that adding 500 milliseconds to search page generation dropped traffic by 20%. The Aberdeen Group put the cost of a single second of delay at roughly 7% of conversions. Different studies, same direction.
| Source (dated) | Finding |
|---|---|
| Amazon (widely reported, 2006 to 2012) | Every 100ms of added latency measurably reduced sales. |
| Google (Marissa Mayer, 2006) | Adding 500ms to page generation cut search traffic by 20%. |
| Aberdeen Group (2008) | A 1 second delay was associated with a 7% drop in conversions. |
| Deloitte and Google (2020) | A 0.1s mobile speed gain lifted retail conversions 8.4%, average order value 9.2%. |
Primary sources linking speed to commercial outcomes. Treat as directional; your own numbers will differ.
Read these as directions, not guarantees
These figures come from specific sites at specific times. The consistent finding is the direction: faster earns more, slower earns less. The size of the effect on your site depends on your traffic, your margins, and how slow you are today. That is why the next section shows you how to measure it for yourself.
How much does bounce rate rise as load time grows?
Bounce rate climbs steeply, and the climb accelerates. Google and SOASTA analysed millions of mobile pages in 2017. As full-page load time grew from 1 to 3 seconds, the probability that a visitor bounced rose 32%. At 5 seconds it was up 90%. At 10 seconds it was up 123%. The curve is not linear: the first few seconds are the ones that matter most.
| Load time (1s to ...) | Increase in bounce probability |
|---|---|
| 1 to 3 seconds | +32% |
| 1 to 5 seconds | +90% |
| 1 to 6 seconds | +106% |
| 1 to 10 seconds | +123% |
Google and SOASTA, Think with Google, 2017 (mobile page load vs bounce probability).
Plenty of budget WordPress hosting delivers 2 to 4 second load times under real traffic, and worse when a shared server is busy. That places a large share of your visitors squarely on the steep part of this curve. They leave before your content, products or services ever render. You pay to acquire that traffic; slow pages spend it and get nothing back.
How do I turn a speed number into a revenue number?
Turn speed into revenue with four numbers you already have: monthly revenue, current load time, the seconds you expect to save, and a conversion effect per second. A defensible rule of thumb from the studies above is roughly 2% to 7% of conversions per second saved. Multiply that uplift by your revenue to get the monthly gain, then by twelve for the year.
Worked example. A shop takes £20,000 a month and loads in 4 seconds. You move to hosting that brings it to 2 seconds, saving 2 seconds. At a conservative 3% uplift per second, that is a 6% lift: 0.06 x £20,000 = £1,200 a month, or £14,400 a year. Halve the per-second figure if you want to be cautious; the annual number is still material against a hosting bill.
| Monthly revenue | Uplift at 3%/second (2s saved = 6%) | Extra per month | Extra per year |
|---|---|---|---|
| £5,000 | 6% | £300 | £3,600 |
| £10,000 | 6% | £600 | £7,200 |
| £20,000 | 6% | £1,200 | £14,400 |
| £50,000 | 6% | £3,000 | £36,000 |
Illustrative only. Uses a conservative 3% conversion uplift per second saved, across 2 seconds. Replace with your own conversion data before acting.
Do not guess your load time. Measure it with per-minute monitoring so you see the real figure your visitors experience, including the slow moments a one-off test misses. Our uptime and performance monitoring records response time every minute, so before and after a change you compare real numbers, not hunches.
Why does mobile speed cost more conversions than desktop?
Mobile speed costs more because mobile visitors are slower to load and quicker to leave. Around six in ten web visits now happen on phones, where networks are less predictable and processors are weaker. A page that renders in 2 seconds on a desktop can take 5 to 8 seconds on a mid-range phone over a patchy 4G signal, landing that visit on the steepest part of the bounce curve.
Google and SOASTA (Think with Google, 2017): 53% of mobile visits are abandoned when a page takes longer than 3 seconds to load.
There is a second cost. Google indexes the mobile version of your site first, so weak mobile performance drags your search rankings down as well as your conversion rate. You lose the visitor who arrived, and you lose some of the visitors who would have arrived. For WooCommerce and other online shops, where checkout is the most script-heavy page and often the slowest, that double loss lands exactly where the money is.
How do LCP and Core Web Vitals affect sales and rankings?
Core Web Vitals turn vague speed into three numbers Google measures on real visits, and they feed both experience and rankings. The one that matters most for conversions is Largest Contentful Paint (LCP): the moment your main content appears. Google rates LCP good under 2.5 seconds and poor above 4 seconds. Slow LCP is what a visitor feels as a slow, untrustworthy page.
| Core Web Vital | Measures | Good | Poor |
|---|---|---|---|
| LCP (Largest Contentful Paint) | Time until main content appears | under 2.5s | over 4.0s |
| INP (Interaction to Next Paint) | Responsiveness to taps and clicks | under 200ms | over 500ms |
| CLS (Cumulative Layout Shift) | Visual stability while loading | under 0.1 | over 0.25 |
Google Core Web Vitals thresholds (web.dev).
Core Web Vitals are a ranking signal in Google Search, so the same LCP that decides whether a visitor stays also helps decide whether they find you at all. LCP is heavily influenced by how fast your server returns the first byte of the page. That is why hosting is not a side issue to Core Web Vitals: it is the foundation the score is built on. For a fuller walk-through, see our Core Web Vitals guide for WordPress.
Where does hosting fit, and is it the fastest fix?
Hosting is usually the fastest single fix because server response time sets the floor under everything else. You can compress images, tune caching and trim plugins for weeks, and you should. But if the server is slow to answer, every visitor waits before any of that front-end work begins. That first wait, time to first byte, is the part front-end optimisation cannot remove.
Why shared hosting has a slow floor
On cheap shared hosting your site draws from a pool of PHP workers shared with strangers. When their traffic spikes, your pages queue behind it, so your response time swings with other people's busy moments. On our platform every site runs in its own dedicated container with its own database. Your speed depends on your traffic, not a neighbour's sale.
That dedicated backend is what makes front-end work pay off. With a fast, steady server response, caching and image optimisation land on solid ground instead of papering over a slow floor. Our WordPress performance hosting is built around that isolation, with per-minute monitoring so you catch a regression yourself rather than hearing it from a customer.
If the thing stopping you moving is the move itself, we do assisted migrations for free. We build a preview of your migrated site first, you check it against the live one, and nothing cuts over until you approve. That removes the usual risk of a speed upgrade: switching hosts blind and hoping. See how assisted migration works, or read the wider case in managed versus shared hosting.
Common questions about site speed and conversions
Short, direct answers to the questions people ask most when they suspect speed is costing them sales.
Frequently asked questions
Does website speed really affect conversion rate?
Yes, and the effect is measured repeatedly. Google and SOASTA (2017) found bounce probability rises 32% as load moves from 1 to 3 seconds. Deloitte and Google (2020) measured a 0.1 second gain lifting retail conversions 8.4%. Faster pages convert better; slower pages lose people before they act.
How fast should my website load for good conversions?
Aim for your main content to appear in under 2.5 seconds, which is Google's threshold for a good Largest Contentful Paint. Anything over 4 seconds is rated poor and sits on the steep part of the bounce curve. On mobile, where most visits happen, hitting under 2.5 seconds matters most because phones and mobile networks are slower.
How do I calculate the revenue impact of a slow website?
Take your monthly revenue and the seconds you can save, then apply a conversion uplift of roughly 2% to 7% per second from the published studies. For example, saving 2 seconds at a conservative 3% per second is a 6% lift, which is £1,200 a month on £20,000 of revenue. Use your own conversion data to refine it.
Is upgrading hosting the fastest way to speed up my site?
Often, yes. Server response time is the floor under every front-end optimisation, so if your host is slow, caching and image work cannot fully fix it. Moving to hosting where your site has its own dedicated container and database removes the shared-server slowdowns that make cheap hosting unpredictable.
Does site speed affect Google rankings as well as conversions?
Yes. Core Web Vitals, including Largest Contentful Paint, are a ranking signal in Google Search, and Google indexes your mobile version first. A slow site therefore loses conversions from visitors who arrive and loses ranking, which reduces the visitors who arrive at all.
About G7Cloud Engineering
Articles written by the engineers who build and run G7Cloud: UK managed hosting and the AI Website Builder. We write about what we operate every day: containers, backups, databases, and the small-business websites that run on them.
More about G7Cloud →