Nearly all landing page advice was written for lead generation, where success is a filled-in form. Ecommerce landing pages have a different job, and most of the rules invert.
Search for landing page best practices and you'll get advice built for software companies.
Remove the navigation. One call to action. Keep the form above the fold. Match the message to the ad. All of that comes from lead generation, where the conversion is somebody typing an email address into a box and the entire page exists to make that one action inevitable.
Ecommerce has a different job. The conversion is somebody spending money on a physical object they can't touch, which they'll have to wait for, and which they might have to send back. The rules don't transfer cleanly and several of them invert.
What's different
The decision needs more information, not less. A lead-gen page removes everything that isn't the form, because any extra information is a reason to hesitate. A shopper deciding on a $120 jacket needs the fabric, the fit, the delivery date and the returns policy. Stripping the page doesn't reduce friction. It removes the answers.
People will leave and come back. In lead gen that's a lost conversion. In ecommerce it's normal behavior, and pages that trap people tend to lose them. Removing your navigation stops the shopper checking your shipping policy, so instead they check whether the product is available somewhere they already trust.
The price is the headline. No lead-gen page has to survive the reader knowing the cost up front. Yours does, on the first screen, on a phone.
Social proof does different work. A logo strip of customers means something in B2B. In ecommerce, reviews on the specific product matter and everything else is decoration.
The five things that actually carry an ecommerce landing page
The product is visible immediately, at the right scale. Not a lifestyle banner with the product implied somewhere in it. On mobile, the first screen should contain the thing and its price.
The objection is answered before it's asked. Every product has one thing standing between interest and purchase. Fit, compatibility, durability, whether it's the real version. The page either answers it or the shopper leaves to find out.
Shipping and returns are visible without hunting. Not in the footer. Near the price, where the cost calculation happens.
One primary action, thumb-reachable, persistent. Add to cart, reachable one-handed, still reachable after scrolling.
Everything else is in service of the decision. If a section doesn't answer a question a buyer actually has, it's making the page longer for no return.
Where the traffic came from changes the page
This is the part that gets skipped, and it's the reason generic templates disappoint.
Paid social. The visitor wasn't looking for you. They were scrolling and something caught them. They need the most context and the most reassurance, because they arrived with no intent and no knowledge of your brand. This page should work as an introduction.
Search. They typed a specific thing. They want to see that specific thing immediately and they're comparing. Speed and specificity matter more than storytelling.
Email. They already know you. Skip the introduction entirely and get to the offer. Repeating your brand story to an existing customer wastes the one screen you had.
Retargeting. They've been here. Something stopped them. The most useful page addresses the thing that stopped them, which usually means shipping cost, a sizing question or a price they wanted to think about.
Same product, four different first screens. Most stores send all four to the same page and then compare the channels on conversion rate, which mostly measures how well a single page happens to suit four different situations.
Judge it on the right number
Add-to-cart rate across ecommerce runs around 6% at the median, and swings from under 2% in luxury and jewelry to 17% in beauty. Median mobile session conversion is about 1.2%.
Those are the numbers a landing page moves. Not bounce rate, which in GA4 now means the inverse of engaged sessions and confuses everybody, and not time on page, which rewards confusion.
Measure add-to-cart rate for the page, and revenue per session for the channel that feeds it. If add-to-cart rises and revenue per session doesn't, you pulled in people who add and don't buy, which is a different problem and usually a price or shipping one.
That loop is what we built Qosmic to run. The audit finds which page is failing and at which check, then we build the variant and adjudicate it against real revenue, both halves, so the new landing page comes back with a number rather than a preference.
You can run it with a page builder, a testing tool and somebody who owns it weekly. That works. Plenty of good teams do exactly that.
The short version
- Landing page advice from B2B assumes the goal is a form. Yours is a purchase.
- More information, not less. Removing content removes answers.
- Don't trap people by stripping navigation. They'll leave the site instead of the page.
- Build a different first screen per traffic source. The four audiences are not the same.
- Judge on add-to-cart rate and revenue per session, not bounce rate.
