What checkout completion rate measures
Count the people who started checkout, count the ones who paid, and divide the second by the first. It deliberately ignores everyone earlier in the journey. Browsers, comparers and people who added an item to think about it are all excluded, so what is left is the group who had decided to buy and then met your checkout.
That makes it the most actionable metric in an online shop. A visitor who leaves a product page may simply have been the wrong visitor. A visitor who leaves after entering their address was the right visitor, and something you built lost them.
The measurement rests on the begin_checkout event firing at a consistent point. If it fires on the cart page for some visitors and on the address step for others, the denominator changes meaning between templates and the rate becomes unreadable. Fix that before drawing any conclusions from the number.
Why checkout completion rate matters
The money has already been spent by this point. Every abandoned checkout represents an ad click you paid for, a page that worked, a product that appealed, and then a form that did not. Recovering some of that group is nearly always cheaper than buying more traffic to replace them, because there is no media cost attached to the fix.
It also isolates blame. Read against the add to cart rate, it tells you whether to work on traffic and product pages or on the checkout itself. Those are different jobs with different owners, and shops routinely spend months on the wrong one.
In Nepal there is a specific reason to watch it closely. Card and digital wallet checkouts fail more often than shop owners assume, cash on delivery is expected by a large share of buyers, and a checkout that offers only one payment path will show a weak completion rate that no amount of ad optimisation can lift.
Where checkout completion goes wrong
Surprises cost the most. A delivery charge revealed at the last step, a tax line the customer did not expect, a minimum order they only now discover — each one contradicts something the shop implied earlier, and people leave when the price changes shape.
Forced account creation is the next offender, followed by forms that ask for more than delivery requires. Every extra field is another chance to stop. On a phone, an address form that resets on error, a keyboard that shows letters when the field wants digits, or a payment page that opens in a window the browser blocks will all lose people who genuinely wanted to buy.
The final one is silent failure. Payments decline for real reasons, but a checkout that shows a blank error and no next step converts a recoverable problem into a lost sale.
What to do about it
Complete your own checkout on a mid-range phone, on mobile data, using each payment method you offer, and write down every moment of hesitation. That exercise finds more than most analytics reports do. Then remove one obstacle at a time: guest checkout, delivery cost shown from the cart onward, fewer required fields, the payment options your customers actually use, and clear recovery wording when a payment fails.
Measure by step rather than in aggregate, so you can see which stage loses people, and by device, because the answer is usually different on a phone. Treat the work as ongoing conversion rate optimisation rather than a one-off project, and pair it with a recovery message to the people who left, which is the cheapest sale available to most shops.