SEO guides

Set a cold-load budget for a mobile product page

·

A performance budget gives a team transfer and interaction limits to check before release. This example uses a product page and keeps the assumed cold transfer separate from the repeat-visit estimate.

Core Web Vitals Budget Planner

Check the cold and repeat allowances

Choose Load product page example and Generate budget plan. Constrained mobile supplies 1,600 kbps. Moderate complexity and the device profile reserve 900 ms. The 2,500 ms LCP goal leaves 1,600 ms available, yielding a modeled 320 KB cold transfer envelope.

The product-page plan allocates 64 KB to first-party JavaScript, 38 KB to stylesheets, 137 KB to first-view images, 22 KB to fonts and 25 KB to third-party scripts. It leaves 34 KB unallocated. At the assumed 80% repeat cache hit rate, the modeled repeat total is 92 KB.

Google Web Vitals thresholds and measurement scopeGoogle introduction to performance budgets

Compare the supplied readings with the targets

The example includes field p75 readings labeled by the user. LCP is 2,700 ms, which exceeds its target by 200 ms. INP is 180 ms against 200 ms. CLS is 0.08 against 0.1. Supply observations from the same page and device scope when comparing a release.

The interaction-work allowance is 88 ms. The supplied 96 ms exceeds it. This work estimate and the measured INP answer different questions. INP also includes input and presentation delay.

Google Web Vitals thresholds and measurement scopeInteraction to Next Paint definition

Use the exported limits before release

Open Compare supplied measurements and change first-party JavaScript from 72 KB to 60 KB. Generate again. The original reading exceeded the 64 KB allowance by 8 KB. The new supplied figure is 4 KB under it.

Download the JSON plan with its assumptions and compare a fresh capture against the same scope. Cold byte limits and the 75th-percentile vitals are separate checks. Reserve space for first-view images and review font shifts alongside the CLS reading.

Google introduction to performance budgetsGoogle Web Vitals thresholds and measurement scope

Common questions

Why is the first-party script cap lower than its transfer allocation?

The scenario applies a second planning cap based on the INP target and the selected device and complexity. Use it as a guardrail for script work. A measured INP reading includes the full interaction.

What gets counted twice if I include third-party JavaScript in both budgets?

Separate first-party JavaScript from third-party scripts when entering supplied transfer sizes, so each script has one category.

Sources

Put it into practice.

Related guides