Your most valuable landing page is rarely your most visited one, and the GA4 interface makes that hard to see. The export lets you rank landing pages by the revenue and conversions they actually start, not just the traffic they receive. Here is how to find them.

What a landing page is in the export

A landing page is the first page a user sees in a session, and the export does not label it for you. You derive it by finding, within each session, the page_view with the earliest event_timestamp. Everything else in this analysis hangs off that one definition, so it has to be right.

Step 1: find each session's landing page

Group events into sessions, then pick the first page_location per session:

WITH sessions AS (
  SELECT
    CONCAT(user_pseudo_id,
      (SELECT value.int_value FROM UNNEST(event_params) WHERE key='ga_session_id')) AS session_key,
    ARRAY_AGG(
      (SELECT value.string_value FROM UNNEST(event_params) WHERE key='page_location')
      ORDER BY event_timestamp LIMIT 1
    )[OFFSET(0)] AS landing_page
  FROM `project.analytics_123456789.events_*`
  WHERE _TABLE_SUFFIX BETWEEN '20260101' AND '20260131'
    AND event_name = 'page_view'
  GROUP BY session_key
)
SELECT landing_page, COUNT(*) AS sessions
FROM sessions
GROUP BY landing_page
ORDER BY sessions DESC

The ARRAY_AGG with ORDER BY event_timestamp LIMIT 1 grabs the earliest page per session, which is the landing page. That alone gives you sessions by landing page, the traffic view.

Step 2: attach value, not just traffic

Traffic is the least interesting part. To rank by value, join each session's landing page to whether it converted and how much revenue it produced:

WITH events AS (
  SELECT
    CONCAT(user_pseudo_id,
      (SELECT value.int_value FROM UNNEST(event_params) WHERE key='ga_session_id')) AS session_key,
    event_name, event_timestamp,
    (SELECT value.string_value FROM UNNEST(event_params) WHERE key='page_location') AS page,
    ecommerce.purchase_revenue AS revenue
  FROM `project.analytics_123456789.events_*`
  WHERE _TABLE_SUFFIX BETWEEN '20260101' AND '20260131'
),
landing AS (
  SELECT session_key,
    ARRAY_AGG(page ORDER BY event_timestamp LIMIT 1)[OFFSET(0)] AS landing_page
  FROM events WHERE page IS NOT NULL GROUP BY session_key
)
SELECT
  l.landing_page,
  COUNT(DISTINCT e.session_key)                       AS sessions,
  COUNTIF(e.event_name = 'purchase')                  AS purchases,
  ROUND(SUM(e.revenue), 2)                            AS revenue
FROM landing l
JOIN events e USING (session_key)
GROUP BY l.landing_page
ORDER BY revenue DESC

Read it by value per session

A page with fewer sessions but high revenue per session is often more valuable than a high-traffic page that converts no one. Add a revenue-per-session column to surface that:

value per session = total revenue รท sessions

Sorting by that ratio, rather than raw revenue or raw sessions, is what reveals the underrated pages worth more investment.

Common traps

A few things distort landing-page value if you let them:

  • Query strings splitting one page into many URLs, which you fix by stripping parameters from page_location.
  • Counting revenue on the landing page's session, not the page itself, which is correct here but worth stating so no one misreads it.
  • Ignoring the suffix filter and scanning months you do not need.

Normalize the URLs first

Landing-page analysis falls apart when one page appears as a dozen URLs because of query parameters and fragments. Strip them before you group, with REGEXP_REPLACE on page_location, so /pricing, /pricing with a utm_source, and /pricing with a fragment collapse into a single row. Without that step your most valuable page is scattered across variants and hidden in plain sight.

Slice by channel to find the real winners

A landing page that converts cold paid traffic is worth more than one that only converts warm email clicks, so pair landing page with channel. Grouping by both reveals which entry points turn expensive traffic into revenue, which is exactly where more budget pays off. You can also add new-versus-returning as a dimension, since a page that acquires first-time buyers is strategically different from one that mostly serves repeat customers.

Your most valuable landing page hides behind your most visited one, and the export is what pulls it into the light. Define the landing page as the first page of a session, attach conversions and revenue to it, and rank by value per session rather than traffic. Do that and you stop pouring budget into busy pages that convert no one and start feeding the quiet pages that actually pay.

Want a stronger data analyst role or a raise? Grab the FREE Product Analyst Playbook and get the exact roadmap to your next offer.