Most stores have one number for their load time. One LCP, one p75, one traffic light in a dashboard. It is green, so everything is fine.
The trouble with that one number is that it is decided by the pages with the least to do with your revenue. Category pages, product pages, the blog, the home page. That is the bulk of your pageviews, so it is the bulk of your average.
Your checkout is a footnote in there. It has few views, so it barely moves the p75. If it is slow, you will not see it. You will see green.
Why seconds are expensive in that one place
A visitor on a category page is browsing. If it gets too slow, what you lose is a browser, and you have plenty of those.
A visitor in the checkout has already decided. They searched, compared, added to the cart and got as far as typing an address. Every second they wait there risks a sale that was effectively already yours. The same second is worth a multiple at that point.
The Deloitte/Google study Milliseconds Make Millions (2020, 37 brands, over 30 million sessions) found a correlation in retail between 0.1 seconds less load time and 8.4 percent more conversions. That is an observed correlation, not a guaranteed effect, and anyone promising you otherwise is selling something. But the direction is clear enough to warrant the question: do you actually know how fast your checkout is?
Three steps instead of one number
fastmon marks every pageview with the step it sits on: cart, checkout or success. Pageviews outside the funnel keep an empty step rather than being given a fourth value for the sake of completeness.
The assignment happens at ingest, from the cleaned page URL, not from an extra field in the tracker. The script on your site does not grow by a single byte for it. Built-in patterns cover Shopware 5 and 6, WooCommerce with English and German slugs, Magento 2, PrestaShop, JTL Shop 5 and Shopify. Language and store prefixes are accounted for, so /de/warenkorb and /fr/cart count like their root variants.
The order of checks is fixed, and for a reason: success first, then cart, then checkout. Confirmation pages on Shopware and Magento often live under the checkout path, /checkout/finish for instance. Check in the wrong order and every purchase is counted as a checkout view.
So instead of one traffic light you get three: the load time of your cart page, of your checkout and of your thank-you page, each at the p75 of real customers on real devices. And usually exactly one of them is the problem.
The Shopify blind spot
If you are on Shopify, one thing is worth knowing. The Shopify checkout runs on Shopify’s own infrastructure, where your theme script never loads. The tracker on your storefront therefore ends at the cart. Left alone, your funnel shows neither checkout nor purchase, and that is not a fastmon quirk: it applies to every tool that is embedded through the theme.
You close it with a custom pixel that reports three events: checkout reached, payment info submitted, order placed. After that the funnel runs to the end again.
Why there is no abandonment rate here
Every analytics tool shows you a cart abandonment rate. We do not, and that is deliberate.
Every number in the funnel counts something observed: a visitor loaded a cart page. A visitor loaded a thank-you page. A metric called “abandoned” would instead be a claim built on a missing observation. The visitor who did not load a thank-you page may have bought on another device. Or after the daily rotation of how we link pageviews together. Or on a confirmation page that carries no tracker.
An abandonment rate would be a figure that looks exactly like a measurement while being an estimate. We would rather show you how many visitors reached each step and let you interpret the gap yourself. That is less comfortable and it is more honest.
What to do with it
Look at the p75 of your three steps side by side, not at the one number across everything. If the checkout is markedly slower than the rest of your store, you have found where your seconds cost the most. And because each step carries its own Server-Timing values and its own network waterfall, you then also see what is behind it: your server, a third-party script in the payment step, or an image nobody compressed.
The funnel is on by default. If your store has unusual URLs, you set your own patterns per step, and new pageviews are marked accordingly from that moment on.