The image formats worth the conversion cost
AVIF is smaller, slower to encode, and not always the right answer. Here is where each format earns its place.
LCP, INP and CLS: what each one actually measures, which of them your users notice, and how far a single fix moves them.
AVIF is smaller, slower to encode, and not always the right answer. Here is where each format earns its place.
Teams optimise the hero image for a fortnight and discover the LCP element was a paragraph of text the whole time.
A lazy attribute on the hero image delays the one thing the metric is measuring.
Lab numbers start arguments. Field numbers end them, because nobody can claim the lab was configured wrong.
It is almost always one of four things, and only one of them is your fault.
Stop arguing that it matters in general. Show what it costs here, on this page, this quarter.
Three options, in descending order of how much the marketing team will like them.
Four numbers, one trend line each, and a decision about whether to move the tripwire. That is the whole meeting.