A developer or agency mentions LCP, INP and CLS, and the conversation moves on as if everyone in the room understands what those three things measure. Most people nod along, quietly hoping nobody asks them to explain it back. Here's what they actually mean, in language that doesn't require a technical background.
LCP: Largest Contentful Paint
This measures how long it takes for the biggest visible thing on the page, usually a hero image, a heading, or a large block of text, to actually appear on screen. It's the closest metric to answering the question a visitor is unconsciously asking: is this page ready yet, or am I still waiting? Google considers under 2.5 seconds good, and anything over 4 seconds poor.
INP: Interaction to Next Paint
This measures how quickly the page responds after someone clicks a button, taps a link, or fills in a form field. A page can load quickly and still feel broken if every interaction has a noticeable lag before anything happens. INP replaced an older metric called First Input Delay in 2024 because it measures responsiveness across an entire visit, not just the first interaction, which turned out to be a more accurate reflection of what actually frustrates people.
CLS: Cumulative Layout Shift
This measures how much content unexpectedly jumps around while a page is loading. It's the metric behind the specific frustration of trying to tap a button just as an ad or image loads above it and the whole page shifts, so the tap lands on the wrong thing entirely. A low CLS score means the page stays visually stable as it loads, even before every element has finished appearing.
Why these three specifically, and not something simpler like a single speed score
A single overall score, however convenient, tends to hide exactly which part of the experience is actually broken, which is exactly why it's worth resisting the temptation to optimise for the headline number alone. A page can score reasonably overall while still being frustrating in one specific, fixable way, slow initial load but responsive once loaded, for instance, or fast to load but janky to interact with. Google's own documentation on this explains the reasoning behind splitting real-world experience into these three distinct signals rather than collapsing it into one number.
What to actually ask a developer or agency about this
Not "what's our PageSpeed score" alone, since that number in isolation doesn't say much on its own. Ask which of the three metrics is currently weakest, and what the plan is to improve that specific one, since a targeted fix to the actual weak point is usually far more effective than a general "make it faster" request with no specific target. This is the level of detail every web design project should be reporting on, not just handing over a single number at the end.
Why these metrics keep changing
INP replacing First Input Delay in 2024 is a reminder that Core Web Vitals aren't fixed forever, Google refines them as it learns more about what actually correlates with a good real-world experience. This isn't a reason to ignore them until they stabilise, the underlying goal, a page that loads quickly, responds promptly and doesn't jump around, stays constant even as the specific measurement of it evolves.