Core Web Vitals are three measurements Google takes from real visitors to decide whether your pages felt fast and stable. They are not a grade out of a hundred and they are not measured on your reception computer. They come from actual phones on actual networks, which is where nearly all of your patients are.
You do not need to understand the code behind them. You do need to know what the three numbers mean, because the difference between passing and failing on a dental site is usually one photograph, one widget and one banner. Here is each measurement in plain terms, along with the questions to put to whoever maintains your site.
Largest Contentful Paint asks how long until something useful appears
LCP is a stopwatch on the biggest thing on screen, which on a dental homepage is almost always the hero photo or the video behind your headline. Google treats 2.5 seconds or less as good and anything past 4 seconds as poor, measured on a phone. Until that image lands, the visitor is looking at a blank rectangle where your phone number should be.
The usual cause is boring and easy to fix. Somebody uploaded the photographer's original file, so the phone is downloading a print quality image to show it three inches wide. Resizing and compressing the hero image, and letting the rest of the gallery load only as the visitor scrolls, moves this number more than any hosting upgrade will.
- Save the hero image at the size it is actually displayed
- Drop autoplaying video backgrounds on the homepage
- Lazy load smile gallery photos below the fold
- Serve modern image formats if the site supports them
Interaction to Next Paint asks what happens when a thumb taps Book
INP replaced the older responsiveness measurement in 2024. It watches what happens when somebody taps a menu, opens a treatment accordion or presses your booking button, and it records how long the page took to visibly respond. Google calls 200 milliseconds or less good, and anything over 500 poor. A page can look finished and still be busy, which is why taps feel dead for a second on some practice sites.
Scripts cause this. A chat bubble, a reviews carousel, a cookie banner, a booking widget and a tracking tag from an agency you fired two years ago are all competing for the same phone. Each was installed for a sensible reason and nobody ever removed one. Ask for a list of every third party script on the site, then justify each one by what it has actually produced.
Cumulative Layout Shift is why people tap the wrong thing
CLS measures how much the page jumps around while it loads. You have felt this: you go to tap the phone number, an ad or a banner drops in above it, and you tap something else instead. Google treats 0.1 or less as good. It is the most annoying of the three vitals and usually the cheapest to fix.
On practice sites, layout shift comes from images without dimensions set, fonts that swap after the text has already drawn, and pop ups or promotion bars that appear a second or two after everything else. Reserving the space in advance solves nearly all of it, and it is a small job for a competent developer.
- Set width and height on every image so space is reserved
- Load a fallback font that is close in size to the final one
- Reserve space for any banner, alert bar or cookie notice
- Never insert a pop up above content that is already on screen
The score that counts comes from real visitors, and it lags
There are two kinds of data here and confusing them causes a lot of pointless arguments. Lab data is what you get from running a test tool right now, which is useful for diagnosis. Field data is what Google collects from Chrome users over a rolling 28 day window at the 75th percentile, meaning three quarters of your visitors have to have the good experience for the page to pass.
Two things follow from that. First, a fix made today will not show up in the field numbers for weeks, so do not panic on day three. Second, a site with very few visitors may not have enough data to report at all, in which case use lab tests and your own phone. Turn wifi off, load your emergency page, and count.
Passing is a tie breaker, not a ranking strategy
Be clear about what these numbers buy you. Page experience is one signal among many, and it will not lift a thin page above a practice with better reviews, closer proximity and pages that answer the question properly. Where it matters is between two practices that are otherwise similar, and in the plain fact that people leave slow pages before they ever call you.
That second effect is worth more than the ranking argument. A patient with a broken tooth gives you a couple of seconds. If you want to know whether speed is your biggest problem or your fifth biggest, a free leak check will show you where inquiries are actually falling out of the process before you spend anything on the site.
What to ask a web developer, and what to ignore
Send three URLs, your homepage, your main treatment page and your contact page, and ask for the current mobile LCP, INP and CLS from field data, a list of what is causing each failure, a fixed price and a date. Ask what will change on the page, in words you can understand. Vague talk about optimization without named causes is a sign to get a second quote.
One more thing worth knowing while site work is in progress. A claimed directory page can carry your details reliably in the meantime, and claiming your listing in the Dent-OX directory is free, puts your services, hours and fees in your own words, emails patient requests straight to your front desk, and gets the page indexed by Google, which unclaimed records are not.
The short version
- LCP is how fast the main image appears, and the hero photo is usually to blame.
- INP is how quickly a tap gets a response, and extra scripts are usually to blame.
- CLS is the page jumping while it loads, fixed by reserving space in advance.
- Field data updates over 28 days, so a fix will not show immediately.
- Give a developer three URLs and ask for causes, a price and a date.
How many patients is your practice missing?
The free New Patient Leak Check shows you where inquiries are slipping away, who is outranking you locally, and what to fix first. It costs nothing and it is yours to keep.
Already listed, but not claimed? Your practice is in our national directory, built from the federal registry. Claiming it is free, puts your services, hours and fees in your own words, sends patient requests to your front desk, and gets the page indexed. Claim your listing.
Frequently asked
What are good Core Web Vitals numbers for a dental website?
Google treats Largest Contentful Paint of 2.5 seconds or less, Interaction to Next Paint of 200 milliseconds or less, and Cumulative Layout Shift of 0.1 or less as good, measured on mobile. Passing all three matters less than the plain experience: a patient in pain should see your phone number in about two seconds.
Will fixing Core Web Vitals improve our rankings?
It can help, but it is one signal among many and it will not outweigh relevance, proximity or reviews. Expect it to act as a tie breaker between similar practices. The larger benefit is that fewer visitors give up before the page loads, which shows up as calls rather than as a position change.
Do we need a new website to pass?
Usually not. Most practice sites fail because of oversized images, too many third party scripts and elements with no reserved space. Those are fixable on the site you have. A rebuild only makes sense when the site is also hard to update, poor on mobile, or built on a platform nobody supports.
Why did our score change when we did nothing?
Field data is a rolling 28 day sample of real visitors, so it moves as your traffic mix changes. A busy month of mobile visitors on weak connections can push numbers down without a single change to the site. Look at the trend across several weeks rather than reacting to one reading.