Marketing & GrowthFree
Landing Page Funnel Drop-off Review
Split each funnel step by the page visitors were on, read those pages, and get page fixes ranked by how many visitors each problem costs.
Paste the funnel report from Google Analytics or any other tool, and list the pages your ads or emails send people to. River splits every step of the funnel by the page visitors were on. Then it opens each page and records what a visitor meets: how many form fields and how many are required. It notes how far down the first field sits on a phone, and anything asking for information people do not have to hand. Findings are ranked by how many visitors each problem loses, with the page and the fix named.
A funnel report gives you one drop rate per step, and that rate averages every page behind it. Take an illustrative van insurer with fourteen landing pages. One page starts a quote for 52.4% of visitors and another, on the same ad group with the same offer, for 8.1%. Pooled with the others, the step reads 44.5%, which describes no page at all. Most guides stop at the percentage. This review opens the pages, because a percentage cannot be fixed and a nine-field form can.
It is for paid search and growth marketers, ecommerce teams and agencies who can see a step leaking but cannot say which page or what on it. It works from a step-level report alone and gets much sharper with page and device breakdowns. When the pages are fine but slow, use the Core Web Vitals triage. When the copy is the problem rather than the form, the value proposition and page copy tool rewrites it.
Why a step-level drop rate can't be fixed
Continue the illustrative insurer. Page A puts a registration number box at the top of the page and 52.4% of its visitors start a quote. Page B hides its form behind a button, then asks for nine fields, five required, including an existing policy number most people do not know. It starts 8.1%. Same traffic, same offer, more than six times the rate. Nothing in the funnel report pointed at page B, because its rate was averaged away with thirteen other pages.
Page B had 21,940 visitors. At page A's rate it would start about 11,500 quotes instead of about 1,800. At the account's own rate of 3.94% from quote to policy, that is roughly 383 more policies, or about £79,000 of first-year margin at £206 each, with no extra ad spend. Treat that as a ceiling rather than a forecast. It assumes page B can match a sibling that already works, and the arithmetic should always be shown.
Two analytics settings change your numbers before any page does. In Google Analytics, funnels can be open or closed, and in an open funnel users can enter at any step, so people resuming a saved quote inflate later steps. And data thresholds can withhold rows in some demographic breakdowns, so segment totals may not add up to the overall figure. Check both settings before you trust any comparison between steps or segments.
How it works
Share the report
Paste or attach the funnel report, and list the pages your traffic lands on.
River splits each step
Every step is broken out by page and device so the spread underneath shows.
River reads the pages
Each page with real traffic is opened and its form and layout recorded as a visitor sees them.
Get ranked fixes
Page changes ordered by visitors lost, each written so a developer can pick it up.
What you get
- Every funnel step split by landing page and device before any rate is called a problem
- Each page read for form length, required fields, form placement and how far down it starts
- Fields that ask for things visitors rarely have to hand, such as account or policy numbers
- Problems ranked by visitors lost, with the percentage beside it, because the orders often differ
- The ad's promise checked against the page, so a broken promise is counted in visitors
- The value of each fix estimated from a better page on your own account, with the arithmetic
Common questions
Why not just fix the step with the worst drop rate?
Because the worst rate is often a small step. In the illustrative insurer, quote to purchase lost 84% of people but far fewer visitors than the landing page lost at 55.5%. The first is mostly a price question. The second is a form a marketer can change this month. Ranking by visitors lost puts the fixable problem first.
What does reading the page add?
Something a developer can act on. A step that loses 55% is not a ticket. A form with nine fields, five required, starting two screens down on a phone and hidden behind a button is a ticket. Both describe the same visitors, and only the second gets built. River records the same facts for every page, including the ones that look fine.
Does this need Google Analytics 4?
No. Any funnel report with step names, visitor or session counts and, ideally, the landing page for each step works, including Adobe Analytics, Amplitude or a spreadsheet. GA4 is the common case, so its open funnel and data threshold settings are checked by name. The page reading works whatever analytics tool you use.
What if my report has no landing page breakdown?
River works from the step totals and the pages you list, reads those pages, and says plainly that it cannot yet show the spread between pages. It then tells you which breakdown to add in your analytics tool, usually landing page and device, so the next run can rank pages by visitors lost.
Is this the same as Google Ads landing page experience?
No. Google says Quality Score is a diagnostic tool, not an input to the ad auction, and its landing page rating compares you with other advertisers as above or below average. That rating will not tell you a form asks for nine fields. This review works from your own visitor counts and the pages themselves.
What if the problem starts before the page?
It often does, and splitting by page shows it. If ads promise a quote in 90 seconds and two landing pages have no form near the top, that mismatch is a finding counted in visitors. When the ad itself is the problem, the creative performance analysis reviews which messages work.
Sources
Built by
Chandler Supple
Co-founder & CEO, River
Chandler has spent the last six years working in AI, building tools for millions of users and Fortune 100 companies. In 2025, he co-founded River to build a new kind of operating system for AI agents.
Landing Page Funnel Drop-off Review
Describe what you need, add any files you have, and your workspace opens with the work already underway.