Core Web Vitals: what each metric measures and how to fix it
LCP, INP and CLS measure different failures with different causes. Knowing which metric is failing tells you almost exactly where to look.
ZooBieDoo Infotech Engineering TeamRead
Key takeaways
- Match the page to the promise that generated the click.
- Removing friction usually beats adding persuasion.
- Speed and layout stability are conversion features, not technical details.
- Low-traffic sites should prefer qualitative evidence over underpowered tests.
Match the promise that brought them
The most common cause of a poor conversion rate is a mismatch between the message someone clicked and the page they arrive on. If the advertisement offers a specific thing, the page must lead with that specific thing.
This applies equally to organic search. A page ranking for a narrow query should answer that query above the fold rather than opening with a general company introduction.
Remove friction rather than adding persuasion
Teams reach for more copy, more testimonials and more urgency. It is usually more productive to remove reasons not to act: unclear pricing, an unexplained form field, a required account before a quote, or a missing answer to an obvious question.
- Ask only for information you genuinely need at this step.
- Explain what happens after submission and how quickly.
- Answer the two or three objections that come up on sales calls.
- Make the primary action unambiguous and repeated at a sensible interval.
Speed and stability are conversion features
A page that takes several seconds to become usable on a phone loses visitors before the copy has a chance to work. Layout shift that moves the button as someone reaches for it is worse still.
Performance work is often the highest-return conversion work available, and it is measurable in a way that copy changes are not.
Test honestly
Small sites rarely have the traffic for statistically meaningful A/B testing, and running an underpowered test produces confident conclusions from noise.
Where volume is low, prefer qualitative evidence: watch people use the page, read the questions that arrive by email, and look at where forms are abandoned. These give clearer direction than a test that cannot reach significance.
What to check before spending on traffic
Paying to send traffic to a page that does not convert is an expensive way to discover a design problem. A short checklist before any campaign spend is worthwhile.
- Does the page load and become usable quickly on a mid-range phone?
- Is the offer identical to the wording in the advertisement?
- Is the form as short as the task genuinely requires?
- Is the primary action visible without scrolling on a phone?
- Is analytics recording the conversion correctly — verified, not assumed?
That last one causes more wasted spend than any design decision, because a campaign that appears to fail may simply be unmeasured.
Frequently asked questions
Every field costs completions. Ask for what you need to respond usefully and collect the rest in the follow-up conversation. Name, contact method and a description of the problem is enough for most enquiries.
For paid campaign landing pages, reducing navigation often helps by removing exits. For organic landing pages, keeping navigation is usually better because visitors may want to research more widely before acting.
As long as it takes to answer the questions a buyer has, and no longer. Considered purchases need more; simple ones need less.
