How to find where visitors drop off on your website, step by step
Drop-off has three levels: which step of the journey, which page, and which spot on that page. Find them in that order. Each one narrows down where to look next.
Quick answers
How do I find where visitors drop off?
Write down the steps from first visit to goal, build a funnel for those steps in GA4 or Clarity, and find the step with the biggest loss. Then look at that page's scroll depth and recordings to find the spot where people stop.
Which free tools can do this?
GA4's funnel exploration shows which step loses visitors, with up to 10 steps and a breakdown by device. Microsoft Clarity is free and adds scroll depth, recordings, and dead and rage clicks on the page itself.
Where does Cortality fit?
Cortality measures how real visitors use every element of the page: what they see, what they skip, and where they scroll, tap and stop. You get a ranked Fix List with the exact change for each spot, and a re-measure that proves whether it worked. Start your free week.
Step 1: Write down the path
Before opening any tool, list the steps a visitor takes from arriving to doing the thing you want. Keep it short. For example:
- Lead-gen site: landing page → contact page → form started → form submitted
- Online store: product page → added to cart → checkout started → purchase
- SaaS site: homepage → pricing page → signup page → account created
Each step needs something you can track: a page view or an event. If a step has no event (for example "form started"), add one before you go further. You can't find a drop you aren't recording.
Step 2: Find the step that loses the most people
In GA4, open Explore and pick the Funnel exploration template. Add your steps in order. Google's help page says you can define up to 10 steps, choose an open or closed funnel, and break each step down by a dimension such as device category (GA4 funnel exploration help, checked October 29, 2026).
Two settings to get right:
- Closed funnel if visitors must start at step 1, which suits most landing pages. Open funnel if people can enter at any step, for example straight onto pricing from search.
- Breakdown by device category. Many drops are phone-only. If phones lose far more people than desktop at one step, you've already narrowed the problem down a lot.
Clarity also supports funnels, and it's free. Microsoft calls it "a free service forever" with no traffic limits (Clarity FAQ, checked October 29, 2026).
Pick the step with the biggest loss that's closest to the goal. That's the page to study.
Step 3: Find the spot on that page
A funnel tells you which page loses people, not where on the page. For that:
- Scroll depth. Where do most visitors stop? Clarity lets you filter recordings by page scroll depth to find visitors who leave before seeing the whole page.
- Recordings. Watch 10 to 20 sessions of people who left at this step. Look for the moment they hesitate, scroll back up, or leave.
- Dead and rage clicks. Clarity flags dead clicks (clicks that get no response) and rage clicks (repeated clicks in one spot). Both point at something broken or confusing.
- Form fields. If the step is a form, find the field where people quit. Form abandonment analysis covers the patterns.
Write down what you see as a specific spot: "visitors stop at the shipping cost table" is useful, "people don't like the page" isn't.
Step 4: Match the spot to a likely cause
| Where they stop | Common cause | First thing to try |
|---|---|---|
| Within the first screen | The page doesn't match the ad or search that brought them | Make the headline repeat what they clicked on |
| Just above the main button | They never see it, or it doesn't stand out | Move the button up or repeat it; see which elements visitors ignore |
| At the price | Price is a surprise, or what's included is unclear | Show price and what's included together |
| At a form field | The field asks too much, too early | Cut or make optional any field you don't need |
| At checkout | Unexpected costs, forced account creation | Show costs earlier; offer guest checkout |
| Only on phones | Layout, tap targets, or overlays covering content | Test the page on a real phone |
Step 5: Change one thing, then measure again
Fix the spot you found, write down the date, and compare the same funnel step for a similar period before and after. Change one thing at a time. If you change three, you won't know which one helped.
On low-traffic sites, a clean A/B test often isn't possible. A before-and-after comparison on the same step is still far better than guessing. CRO without A/B testing covers how to do it honestly.
Where Cortality fits
Funnels show which step loses people. Recordings show individual sessions, and you have to watch enough of them to spot a pattern. Neither tells you which element on the page caused the drop, or whether your fix worked.
Cortality measures how real visitors use every element of the page: what they see, what they skip, and where they scroll, tap and stop. You get a ranked Fix List. Each finding shows what we measured, why it costs you, the exact change to make, and a re-measure that proves whether it worked. It works alongside GA and Clarity rather than replacing them, and it needs one small tag, which takes about 5 minutes to add.
Want a quick look first? The free page scan works from a URL alone. It takes about a minute, no install needed, and it's a prediction from our attention model, not data from your real visitors. The full findings arrive by email.
Every plan starts with a free first week, and you see your first report before any charge. Plans are $99, $249 or $499 a month, and any charge can be refunded within 14 days, no reason needed.
Questions
Is drop-off the same as bounce rate?
No. Bounce rate only covers visitors who leave from the first page without engaging. Drop-off is any point where visitors stop moving toward the goal, which includes people who scroll halfway down a pricing page, start a form and quit, or reach checkout and leave.
Can I find drop-off without GA4?
Yes. Microsoft Clarity is free and shows scroll depth, recordings and funnels. Any tool that records page views and a few key events can show which step loses people. You need some event data either way, so nothing works on a site where you can't add a tag.
How many visitors do I need to trust the numbers?
Enough that a handful of sessions aren't deciding the answer. On small sites, look at a longer date range and focus on the biggest drop, not small differences between steps.
What should I fix first?
The step with the biggest loss that sits closest to the goal. Losing people on the pricing page or the form usually costs more than losing them on a blog post, because they were closer to buying.