Review a landing page by traffic source before changing the design

Your homepage may receive the most visits without being the page that introduces most prospective customers to a particular service. An email can send people straight to an offer, while a search result can bring them into an article. A useful page review starts with the visitor’s entry point and the reason they arrived.

GA4’s landing-page report helps identify those entry points. Combine it with a checked enquiry measurement and a manual review of the page. The report can point to a question worth investigating, but it cannot tell you which headline or design will fix it.

Start with the first page of the session

In Google’s Landing page report guidance, the landing page is associated with the first page view in a session. It can be a product page, an article or another page, rather than the homepage.

Imagine a fictional business visitor enters through /boiler-servicing/, reads a pricing page and submits an enquiry. For this review, the service page is the entry point. The fact that the visitor later saw Contact does not make Contact their landing page.

Choose a stated reporting period, then inspect the relevant page and add Session source / medium as a secondary dimension where the report allows it. Navigation can vary if the property’s reports have been customised. Google also describes a Traffic acquisition route for combining source information with landing pages.

Compare traffic with a similar purpose

A service page receiving local search visitors and an article receiving international social traffic are answering different needs. Their enquiry counts alone do not establish which page is better.

Before comparing two entry pages, write down what is held reasonably constant. This might include the offer, geography, traffic source and reporting dates. Note what differs. If one page was used for a discounted offer, record that rather than crediting the design for the entire difference.

Small numbers need particular care. One extra enquiry can change a percentage sharply when there are few sessions. There is no universal visit count that makes every small-business comparison reliable. Sometimes the honest decision is to collect more relevant observations while fixing obvious faults such as a broken form.

A worked example with consistent units

The following numbers are fictional and illustrate the calculation. They are not results from a campaign or a published benchmark.

• Page A received 120 sessions from the selected source and period.
• Page B received 80 sessions from that same source and period.
• Six of Page A’s sessions included a confirmed-enquiry key event.
• Two of Page B’s sessions included a confirmed-enquiry key event.

The proportion of sessions with an enquiry is 6 divided by 120, or 5%, for Page A. For Page B, 2 divided by 80 is 2.5%.

Those numerators are sessions containing the enquiry action. They are not simply the total number of event occurrences. If one visitor submits twice in a session, two enquiry events do not represent two sessions with enquiries. Use a correctly configured session-based rate or a calculation whose numerator matches its denominator.

Page A has the higher observed rate in this example. That does not prove its layout caused the difference. The visitors, offers or implementation could still differ, and two versus six enquiry sessions leaves substantial uncertainty. Check the actual enquiries before treating either group as commercial success.

Investigate zeros before rewriting the page

If a page shows traffic and no key events, test the desired action. The event may be absent, incorrectly triggered or newly marked as important. Google’s key-event documentation explains that marking an event does not backfill historical reporting.

Compare the form inbox or enquiry system with what Analytics collected during the same period. A measurement fault needs a measurement repair. Changing a headline will not recover missing events.

Inspect the page against the promise in the link

Once the data is usable, open the page on a phone and follow the journey yourself. The following checks are practical editorial suggestions, rather than a Google scoring formula.

1. Match the offer. Read the advertisement, email or search-facing description beside the page. Can a visitor recognise the service or offer they expected?
2. Resolve the immediate questions. Check whether the page explains what is provided, who it suits and any location or eligibility limits that matter.
3. Follow the next step. Find the enquiry or booking action without relying on your knowledge of the site. Check that its wording explains what happens next.
4. Complete the form. Look for unclear required fields, validation messages and a meaningful success response.

Search presentation deserves its own check. Google’s SEO Starter Guide recommends understandable content and descriptive URLs. Those checks help people and search engines understand a page, but they do not guarantee rankings or enquiries.

Record a change you can review later

Choose a specific issue supported by the inspection. For example, an offer page might omit the service area even though the promotion targets local customers. Record the omission, the proposed wording and the date it changes.

Keep the original reporting period and compare later traffic under similar conditions. If the offer, advertising budget and page all change together, describe the result as a combined observation. A before-and-after percentage cannot isolate the effect of one sentence.

Leave a review date and an owner against the change. When the next report arrives, revisit the original question: did visitors from this source find a suitable service and complete a useful enquiry? Keep the answer tied to the observations you actually have.