Website Performance Budget Calculator

Set concrete page-weight and Core Web Vitals targets for your site based on who you are serving and how they connect. Add your current load time and revenue figures to see what a slow site is actually costing you. Useful for briefing developers, prioritising improvements, or making the case for a performance budget to stakeholders.

Explain like I'm 5 (what even is a performance budget?)

Imagine your website is a rucksack. A performance budget is deciding in advance how much it's allowed to weigh before someone's too tired to carry it. Every image, script and font you add is weight. Without a limit, rucksacks (and websites) tend to get heavier over time until someone gives up and leaves. Setting a budget forces you to make deliberate trade-offs.

Calculate

Select your site type and connection, then press Calculate.

Prove it

Budget figures are based on HTTP Archive data and Google's Web Performance working group recommendations, adjusted by site type. Revenue impact uses a conservative 7%/sec conversion loss rate for e-commerce (Portent/Google research), 5% for SaaS, 2% for content sites, applied to seconds above a 2-second baseline. Capped at 80% maximum estimated loss.

Useful? Press Ctrl + D to bookmark.

Why performance budgets matter

The median web page has grown considerably over the past decade. What shipped in a few hundred kilobytes in 2012 often ships as several megabytes today. Most of that growth is not deliberate: it is the accumulation of one reasonable decision after another, made without any constraint on the total.

A performance budget changes that. By agreeing upfront that the JavaScript bundle cannot exceed 300 KB, or that LCP must stay under 2 seconds, you force trade-offs to happen explicitly rather than silently. When a proposed feature would break the budget, you have to decide whether to skip it, optimise it, or consciously accept the cost.

How to use these numbers

The budgets shown here are targets, not guarantees. Start by measuring where you are: run your site through PageSpeed Insights or WebPageTest to see your real page weight and Core Web Vitals scores. Then compare against the targets for your site type and audience. The gap tells you where to focus.

Typical quick wins: compress and convert images to WebP or AVIF, defer non-critical JavaScript, remove unused CSS, self-host fonts. These often cut 30–50% of page weight without changing any functionality.

About the revenue impact estimate

The revenue calculation uses research from Google, Portent and Deloitte that consistently shows conversion rates decline as load time increases. For e-commerce, the effect is largest: roughly 7% conversion loss per extra second above a 2-second baseline. For content sites it is closer to 2%. These are industry averages,your actual numbers depend on your audience, device mix, and what you are selling. Treat the output as a directional estimate for business cases, not an exact forecast.

Related calculators

Performance budgets touch hosting, bandwidth and the commercial case.

Frequently asked questions

What is a performance budget?

A set of agreed limits on your site's size and speed metrics. Defining them upfront means trade-offs are made deliberately, not accidentally. Common budgets: total page weight, JavaScript bundle size, LCP, CLS, INP.

What are Core Web Vitals?

Google's standardised UX measurements: LCP (how fast the main content loads), CLS (how much the page shifts during loading), INP (how quickly the page responds to clicks and taps). Google uses them as a ranking signal.

How does page speed affect conversions?

Research from Google, Portent and Deloitte shows slower pages convert worse. E-commerce sites loading in 1 second convert roughly 3× better than those at 5 seconds. Each extra second typically costs 4–7% of conversions for commercial sites.

What page weight should I aim for?

It depends on your audience. A mobile-first content site should target under 500 KB. A desktop e-commerce site can tolerate more. Define constraints based on your real user data in Google Search Console or Analytics rather than generic benchmarks alone.