On this page
- Why one goal beats one button
- What the first screen has to answer
- How simple should the copy be?
- Where proof and forms belong
- Design for the phone, then check the desktop
- Why the hero image decides your load time
- WordPress checks that catch pages that look finished
- When A/B testing is worth it
- Frequently asked questions
Key takeaways
- Pick one goal and repeat its button at the top, after the persuasive content and at the bottom. A second, different goal competes with the first.
- Unbounce’s 2024 benchmark found pages written at a 5th to 7th grade level converted at 11.1%, against 5.3% for professional-level copy. That is a correlation, not proof that simpler copy causes the lift.
- Don’t lazy-load the hero image. It is usually the largest element on the page, and Google counts a Largest Contentful Paint of 2.5 seconds or less as good.
- In a block theme, a Front Page template overrides Settings > Reading. That is why some homepage changes appear to do nothing.
- A checkout that looks right in the editor proves nothing until the page is assigned under WooCommerce > Settings > Advanced and you have placed a test order on the storefront.
You pay for the traffic. The page looks polished. The leads don’t come. When practitioners audit pages like this, they rarely blame the color palette. They find one of three problems: visitors can’t tell what the offer is, nothing nearby gives them a reason to trust it, or the page is broken in a way nobody saw from the editor.
The first two are design problems. The third is mostly a WordPress problem, and it is the one most best-practice lists leave out. This article covers all three. If you haven’t built the page yet, start with how to build a landing page in WordPress and come back once you have a draft.
Why one goal beats one button
A landing page asks the visitor to take one action. Book a consultation, start a trial, or buy the bundle. If it also offers the newsletter, the blog and a careers link, each one is a decision the visitor has to make before they act.
“One goal” doesn’t mean one button. WordPress.com’s product landing page guide recommends putting a call to action at the top, another after the persuasive content and another at the bottom. That is the same goal three times, placed where people make up their minds. Two different goals on one page compete with each other.
Write the goal down before you touch the editor: “A qualified patient requests a consultation.” That sentence tells you what to measure. Count completed requests, not clicks on a button that opens a calendar.
What the first screen has to answer
Before anyone scrolls, the first screen should say what this is, who it’s for and what to do next. In practice that means four things:
- a headline that names the outcome or the service
- a supporting line with the detail that qualifies the visitor, such as location, price range or availability
- an image of the actual product or result, not a mood photo
- a button that describes what happens when it’s clicked
“Check availability” and “Request a quote” are different promises. Choose the label that matches where the button goes. If a button says “Check availability” and opens a general enquiry form, change the label or the destination. “Learn more” says nothing at all.
The headline also has to continue whatever brought the visitor in. If the ad said “emergency roof repair in Leeds”, that phrase belongs at the top of the page, not three sections down below a general company introduction. Practitioners call this message match. In community threads it’s one of the most common explanations for a page that converts badly.
In the block editor, check the Cover or Group block that holds the hero. A large minimum height, heavy padding or vertical centering can push the button below the first screen on a phone, even when it fits on a laptop.
How simple should the copy be?
Simpler than you think, and as long as the offer needs. Unbounce’s 2024 Conversion Benchmark Report covered more than 41,000 landing pages. Pages written at a 5th to 7th grade reading level converted at 11.1%, compared with 5.3% for professional-level copy. Difficult words, reading time and word count all correlated negatively with conversion.
These are correlations on Unbounce-hosted pages. They aren’t proof that cutting words causes the lift. A newsletter signup needs a headline and a field. An unfamiliar service at a high price needs proof, details and an FAQ that answers the objections people actually raise. Cut words that don’t explain the offer, and keep the ones that do.
The same report puts the median conversion rate across industries at 6.6%, with industry medians ranging from 3.8% for SaaS to 12.3% for events and entertainment. Compare yourself with your own industry and offer, not with the overall figure.
Where proof and forms belong
Specific proof works better than general praise. “They explained our VAT position in one call and onboarding took a week” tells a prospect something. “Great service” doesn’t. Put proof next to the place people hesitate, usually beside the button or the form. A testimonials section at the very bottom doesn’t do that job. Use names, photos and logos only when you have permission.
A form should ask only for what you need to qualify the lead and reply. Keep a field if its answer changes how you follow up. Remove it if it’s there because the template had it. The W3C’s forms tutorial says every control needs a real label, so a placeholder that disappears as the visitor types isn’t enough. Error messages should name the field and say how to fix it.
Then send a real test enquiry from your phone. Check that the right inbox receives it, that the visitor sees a useful confirmation, and that spam protection doesn’t silently reject real submissions. A form that looks perfect but loses enquiries is broken.
Trust also depends on what you forget to remove. Auditors regularly find Lorem ipsum, demo pages left in the sitemap and template phone numbers still linked from the call button on live pages. Read the published page from top to bottom before any traffic arrives.
The same thinking applies to navigation. Remove the journeys that compete with the goal. Keep the information people need before they trust you: contact details, policies and the cart on a shop. Stripping the header and footer completely suits a short paid campaign better than an unfamiliar professional service.
Design for the phone, then check the desktop
In Unbounce’s dataset, 83% of visits came from mobile, while desktop converted 8% better on average. Most of your visitors are probably on a phone, and the phone is where the friction is. Design for it first.
Page builders make this harder than it looks. Elementor’s responsive settings cascade from desktop to tablet to mobile, so a margin set for desktop carries down unless you override it. “Hide on mobile” fixes one screen by removing content from another, and Google indexes the mobile version of your page. Our guide to making a mobile-friendly website covers the theme and builder settings in detail.
Two accessibility checks catch a lot of mobile breakage. WCAG 2.2 requires a contrast ratio of 4.5:1 for normal text and 3:1 for large text. It also requires content to reflow at a width of 320 CSS pixels without losing information or function. Test on a real phone with the keyboard open, and watch for a sticky bar or cookie banner covering the button.
Why the hero image decides your load time
On most landing pages the hero image is the largest element on the first screen, so it sets your Largest Contentful Paint. Google’s Core Web Vitals treat LCP of 2.5 seconds or less as good, along with INP of 200 milliseconds or less and CLS of 0.1 or less, measured at the 75th percentile of real visits.
The most common self-inflicted delay is a “lazy-load all images” setting in an optimization plugin. web.dev’s LCP guide explains that lazy-loading the LCP image delays the moment the browser starts fetching it. Turn off lazy loading for the hero and keep it for images further down. Put the hero in the page HTML as a real image rather than a CSS background, give every image a width and height so the button doesn’t jump while the page loads, and compress the file. Our guide to optimizing pictures in WordPress covers formats and sizes.
Speed shows up in results. Portent’s 2022 study found that a B2B site loading in one second converted three times better than one loading in five seconds. That is an association across a limited set of sites, but it points the same way as everything else.
Measure the live, cached page, not your staging copy. Staging often runs without caching and looks worse than production. Run PageSpeed Insights more than once, and remember that purging the cache slows the site for a while as pages rebuild.
WordPress checks that catch pages that look finished
A page can be well designed and still not work. These checks come from official documentation. Each one fails without warning.
| Check | Where | What goes wrong |
|---|---|---|
| Which template is served | Appearance > Editor > Templates | In a block theme, a Front Page template is used for the front page whatever Settings > Reading says. |
| Header and footer | Elementor page settings | Canvas removes the theme header and footer. Full Width keeps them. Check the published page, not the editor. |
| Old Elementor tutorials | Elementor Pro | Elementor’s separate Landing Page feature is deprecated and unsupported as of Pro 3.27. Build on a regular page instead. |
| Indexing | SEO plugin settings, robots.txt | noindex keeps a page out of Google. Blocking the page in robots.txt can stop Google from seeing that noindex tag. |
| Checkout | WooCommerce > Settings > Advanced | A custom checkout page that isn’t assigned here won’t work as checkout, however good it looks. |
| Shared templates | Site Editor | Editing a shared template, including WooCommerce’s Single Product template, changes every page that uses it. |
Run the checks in this order: confirm which page and template WordPress is actually serving, look at it at phone width while logged out, submit the form or place a test order on the storefront, then measure the live page. Logged-in admins often see different banners and a fresher cache than visitors do. Website layout and design on WordPress explains how global styles and templates affect individual pages.
If you’d rather describe the change in chat and have it made on the live site, SiteSelf handles design changes on existing WordPress sites, though it refuses pages owned by Elementor, Divi or Beaver Builder and tells you why.
When A/B testing is worth it
Test after the basics are right, not instead of them. A clearer headline won’t help much if the form drops submissions. Once the offer, the first screen and the plumbing are sound, test one thing at a time: the headline, the button or the hero image.
Many small sites don’t have the traffic for a meaningful test. In an Unbounce and Ascend2 survey of 264 marketers at businesses with 500 or fewer employees, reported by MarketingProfs, 41% named low traffic as their biggest testing challenge. With limited traffic, don’t declare a winner after a handful of enquiries. Look at form errors, the questions customers ask, session recordings used with proper consent, and whether the traffic was ever a good match for the offer.
Pages also decay after launch. Offers expire, prices change, forms stop reaching anyone and mobile layouts drift after updates. Put a date in the calendar to reread the page and resubmit the form.