If you pay for clicks and send them to your homepage, you are doing it wrong. A homepage serves many visitors with many goals. An ad landing page serves one visitor with one goal, and that gap is where your ad budget leaks.
This guide covers what a landing page for ads must do, how fast it must load, what to put on it, and how to test it without fooling yourself. We run paid media and conversion work for ecommerce and direct-to-consumer brands, so the examples lean that way.
Key takeaways
- A landing page has one job and one action. A homepage has many jobs. Unbounce calls homepages built for exploration and landing pages built for a single campaign or offer. Paid traffic belongs on the second.
- Speed is a conversion lever you can fix this week. Think with Google reports 53% of mobile visitors leave a page that takes longer than three seconds to load.
- Message match matters more than design polish. The ad promise and the first screen must say the same thing.
- Benchmarks are directional. One 2026 roundup puts the median landing page conversion rate near 6.6%, but your category, offer and traffic source will move that a lot.
- Test one change at a time, and only when you have the traffic to read the result.
Table of Contents
- Homepage vs landing page: the difference is intent
- What a landing page for ads must do
- Speed: the maths of the mobile click
- Choosing where to build the page
- Build the page around one ad group, not one brand
- Benchmarks: use them, don’t believe them
- How to test without fooling yourself
- Our perspective: the page is where agencies drop the ball
- A practical build order for this week
- Where we can help
- FAQ
- Sources
Homepage vs landing page: the difference is intent
A homepage is a lobby. It has a navigation bar, a blog link, a careers link, a store locator and six product categories. That works for someone who already knows you and wants to look around.
An ad click is different. The person clicked because of a specific promise: a product, a price, an offer, a problem. Every extra link on the page is another exit from that promise. Unbounce makes the same point: repurposing a homepage for paid traffic adds distractions that erode conversion focus.
The mechanism is simple. Your ad platform charges you per click or per impression. The page decides what that click is worth. Send 1,000 clicks to a page with 12 possible actions and you split intent 12 ways. Send them to a page with one action and you test one outcome.
So what is the difference in practice?
| Homepage | Ad landing page | |
|---|---|---|
| Audience | Everyone | One ad group or audience |
| Goal | Explore, orient | One action: buy, book, sign up |
| Navigation | Full menu | Logo only, or none |
| Message | Who we are | The exact promise in the ad |
| Measured by | Engagement, visits | Conversion rate, cost per acquisition |
(Homepage ≠ campaign page. Calling it a landing page does not make it one.)
When the homepage is fine
There are exceptions, and we will name them. A brand search campaign, where people type your name, can land on the homepage because they came looking for you. A tiny catalogue of one or two products can also work, if the homepage already is the offer. Past that, you are paying extra for the privilege of confusing people.
What a landing page for ads must do
Forget the 40-item checklists. Five jobs matter, in this order.
1. Match the ad
If the ad says “free shipping on swimwear over a set spend”, the first screen says it too, in the same words, with the same product image. Visitors decide in seconds whether they landed in the right place. A mismatch reads as a bait and switch, even when it is an honest mistake.
Why this matters beyond the visitor: on Google Ads, landing page experience feeds into Quality Score in practice, so a weak page can raise what you pay per click as well as lowering what each click returns. Treat that as a cost on both sides of the ledger.
2. Make the offer obvious above the fold
One headline. One line of support. One button. Put the price, the guarantee or the delivery promise near the button, not three scrolls down.
3. Prove it fast
Reviews, ratings, a recognisable press mention, a count of orders. Pick the proof that answers the objection your audience actually has. A skincare buyer worries about skin reaction. A mattress buyer worries about returns. Different proof, different position.
4. Remove exits
Cut the full navigation bar. Cut the footer link farm. Keep the legal links, because you need them. Every link that does not serve the one action competes with it.
5. Load fast on a phone
This one gets its own section, because it is the cheapest win and the most ignored.
Speed: the maths of the mobile click
Think with Google’s Australian mobile speed analysis found the average mobile landing page takes 22 seconds to fully load, and 53% of mobile visitors leave after three seconds. The page does not carry a clear publication date in the extract we reviewed, and mobile networks have improved since, so treat the exact figures as directional. The mechanism has not changed: a person tapping an ad is on a phone, often on a train or a couch, and will not wait.
Here is what that does to your numbers. Say you buy 10,000 mobile clicks. If 53% leave before the page shows anything, 5,300 of them cost you money and never saw your offer. You pay for all 10,000. Your cost per acquisition is calculated on the 4,700 who stayed.
The benchmark to aim for is Google’s own. Core Web Vitals asks for Largest Contentful Paint, the time the biggest visible element takes to appear, within 2.5 seconds, measured at the 75th percentile of page loads. That means three in four real visits should hit it, not your best test on office wifi.
The usual culprits are boring:
- Hero images uploaded at full camera size.
- A video autoplaying above the fold.
- Six tracking and chat scripts loading before the headline.
- A theme built to do everything, loading code for features the page never uses.
Fix the hero image first. It is typically the LCP element, so it is typically the biggest single gain.
Choosing where to build the page
Nobody asks whether you can build a landing page. The real question is whether the tool will let you build a fast, specific one that you can change on a Tuesday without a developer ticket.
There are three broad routes:
- Generalist website builders. Cheap and quick for a portfolio or a solo business. Ecommerce brands often hit limits in load speed and checkout integration once campaigns scale.
- Your ecommerce platform’s native pages. Pages built inside your store share the product catalogue, the cart and the tracking you already trust. For most Shopify brands this is the sensible default. Shopify’s guide defines the landing page as a page with one specific goal, such as a newsletter sign-up, an app download or a purchase.
- Dedicated landing page tools. Useful when marketing needs to publish without engineering. A modern builder lets a campaign page go live in hours rather than sprints, though you still own the speed and the tracking setup.
Our view: pick the route that keeps the page closest to your product data and your analytics. Every extra system is another place for tracking to break and another script to slow the page.
Before you pick a tool, pick the offer. A fast page for a muddled proposition is a fast failure. If you are still sorting positioning, a go-to-market framework for founders can help you pin down who the page is for and what it promises before anyone opens a builder.
What about AI and free builders?
People ask whether they can get a free page or have ChatGPT write one. You can, and for a first draft of copy it is useful. Caveat: a draft is not a tested page. Free builders often add their own branding and limits, and generated copy tends to sound like everyone else’s. Use them to get to version one, then let data decide version two.
Build the page around one ad group, not one brand
The most common mistake we see is one landing page for every ad. It feels efficient. It is the opposite.
Meta and TikTok audiences differ from Google Search audiences. A person searching “waterproof running jacket” has high intent and wants the product now. A person who stopped scrolling on a video ad is curious and needs a reason. Same product, different first screen.
A workable setup for an ecommerce brand:
- Search ads: land on a product or collection page that matches the keyword, with price and delivery visible. Our Google Ads work starts from this match.
- Paid social: land on a page that repeats the hook from the creative. If the ad opens with a customer complaint, the page opens with the fix. See how we think about scaling in our Facebook Ads scaling piece.
- Video ads: land on a page that continues the story, since the viewer already saw the product in use. Our YouTube ads checklist covers the hand-off.
Why segment this far? Meta’s system treats near-identical ads as one ad, so fifteen thumbnails on one script behave like a single ad. The page works the same way. Fifteen ads sending to one generic page behave like one weak offer. Variety only counts when it changes the message the buyer meets.
Benchmarks: use them, don’t believe them
One 2026 roundup of landing page statistics puts the median conversion rate around 6.6% and the top 10% above 11%. Sit with that. Those are pooled figures from many industries and many types of conversion, including free downloads and event sign-ups.
A free ebook converts at a very different rate from a $300 product. The roundup is an aggregator, not a primary study, so we use it as a rough compass only. Nobody can give you an honest universal benchmark for an ecommerce sale from paid social. Anyone quoting one without a category, traffic source and date made it up.
Your benchmark is your own last 90 days. Set it, then beat it.
How to test without fooling yourself
Most A/B tests on landing pages are noise. The reasons are arithmetic.
- Too little traffic. If a page gets 300 visitors a week and converts at 3%, that is nine orders a week. Two versions will not separate in a month. Test bigger swings (offer, layout) rather than button colours.
- Stopping early. A variant that leads on day three often loses by day ten. Decide the run time before you start.
- Changing five things. You learn that the new page won. You learn nothing about why.
If your store sits somewhere between 100,000 and 500,000 monthly visitors, tool choice starts to matter, and we compared the options in our VWO vs Optimizely breakdown for Shopify stores. Below that, simple sequential tests and good judgement beat expensive software.
What if the test does not win? Then you have a result. Keep the control, write down what you learned, and test the next biggest assumption. A flat result on a headline usually says the problem sits in the offer.
Our perspective: the page is where agencies drop the ball
Here is the pattern we see. One team buys the traffic. Another builds the page. A third writes the emails. Each reports its own number, and each number looks fine. Revenue is flat.
The ad team celebrates a low cost per click. The web team ships a page that matches the brand guidelines but not the ad. Nobody owns the gap between click and purchase, and that gap is where the money goes.
We built Thrive to close it. Strategy, creative, paid media and conversion work sit with one team, so the person who writes the ad hook also sees what happens on the page after the click. That is why we judge a landing page by revenue per visitor rather than by how it looks in a review.
Two things we will say plainly. First, a landing page will not rescue a weak offer. If your price, product or proof is off, no layout fixes it. Second, most brands do not need a redesign. They need the first screen to repeat the ad, the page to load in under 2.5 seconds on a phone, and one clear button.
You can see how this plays out in our case studies, including Honeysuckle Swim Co. and Bonaire.
— Harry
A practical build order for this week
If you have one afternoon, work in this order:
- Pull your top three ads by spend. Write down the exact promise each one makes.
- Open each ad on your phone and tap through. Time the page with a stopwatch. Anything past three seconds goes to the top of the fix list.
- Check the first screen against the ad. Same words, same product, same offer.
- Remove the main navigation on the campaign version of the page.
- Compress the hero image and delay any script that is not needed to show the page.
- Set your baseline: conversion rate, revenue per visitor and cost per acquisition over the last 90 days.
- Pick one test and decide the run time before you launch it.
The order matters. Speed and message match are fixes. Testing is optimisation. Do the fixes first.
Where we can help
If your clicks are fine but your sales are not, the problem is usually after the click. Our conversion rate optimisation service covers landing pages, product pages and checkout for ecommerce and DTC brands, and it sits alongside our paid media and creative teams. You can see the full range on our services page, or start from our homepage and tell us what you are running now.
FAQ
Can I use my homepage as a landing page for ads?
You can, but you will usually pay for it. Homepages offer many paths, so ad clicks scatter across them. The exception is a brand-name search campaign, where visitors already want you. For anything else, build a page for the specific offer.
How much should I pay for a landing page?
It depends on the route. A page built in your own ecommerce platform costs your time and a template. A custom design costs agency or freelancer time. We do not publish prices, so we cannot give you a figure here. Judge cost against revenue per visitor, not against the build invoice.
Can ChatGPT build a landing page?
It can draft the copy and a rough structure quickly. It cannot know your audience, your proof or your load speed. Treat the output as version one, then check the first screen against your ad and test it on a phone.
How many landing pages do I need?
One per distinct promise or audience, not one per ad. A search campaign on a product keyword, a paid social campaign built around a customer problem and a video campaign usually need three different first screens. Start with your top-spending ad group and expand from there.
Sources
- Think with Google: Australian mobile page speed benchmarks
- Unbounce: landing page vs homepage vs website
- web.dev: Web Vitals
- Shopify: How to create a landing page
- Guideflow: Best website builder software for 2026
- Involve.me: Landing page statistics (2026)
