Research Note
Fewer Than 4 Percent
What a buyer expects from any website, and what the dealer layer delivers.
Fewer than 4% meet Google’s 2.5 second mobile LCP threshold in a standardized lab test, across 7,700+ verified dealer sites measured. The same population carries an 8.7s median mobile LCP, which is the wait before the main content of a page appears.
Anyone who has stood in a parking lot waiting for a showroom page to load already knows this finding. The part that is easy to miss is what the buyer is comparing it against. They arrive from a search result, a social feed, or a manufacturer’s own site, and all of those render in about a second. The dealer page is the slowest thing they have touched all day, and they have no reason to read that as a dealer problem. They read it as a dead end and go back.
The reason this goes unnoticed is that nobody owns the measurement. A manufacturer measures its campaigns up to the click. A dealer measures whether the phone rang. The page in between belongs to neither report, so it can stay broken for years without appearing in anyone’s numbers as a problem.
What it means for manufacturers
Manufacturers fund demand that lands here. National campaigns, product visualizers, and dealer locators all end at a dealer website, and that page is where brand spend turns into a showroom visit or stops moving. The spend is committed against a layer where fewer than 4% of sites meet that threshold in a standardized lab test.
The dealer network is the last mile of the brand, and it is the stretch a manufacturer has the least visibility into. Every other segment of that path gets instrumented, reviewed, and optimized. The final page gets a link check.
One limitation
This measures page delivery, and it says nothing directly about sales. A slow site still sells, and a fast one can still fail to. What the instrument saw is that the dealer layer is markedly slower than the web its buyers spend the rest of their day on. That is a measurement about delivery and an argument about consequence, and reading it as a quantified revenue loss would claim more than the data supports.