Source: https://measuremy.site/blog/how-to-find-the-biggest-drop-off-in-a-funnel [measuremy.site](https://measuremy.site/) / [Blog](https://measuremy.site/blog) # How to Find the Biggest Drop Off in a Funnel October 11, 2026 ยท 5 min read [Photo by MART PRODUCTION on Pexels](https://www.pexels.com/photo/photo-of-papers-on-table-7605981/) ## Define the funnel in concrete steps A funnel is a sequence of pages or events that a visitor must complete to reach a goal. Before you can find the biggest drop off, write down the exact steps in order. For a signup funnel, that might be: - Landing page view - Click on the sign-up button - Sign-up form submission - Confirmation page view For a purchase, it might be product page, cart, checkout, and thank-you page. Keep the funnel short enough to measure but long enough to find where people leave. If you are not sure which pages matter, start with the smallest set that still represents a real conversion. ## Collect counts for each step You need a count of unique visitors or sessions that reached each step. Total page views are not enough, because one person can refresh a page several times. If your analytics tool lets you build funnels from page paths and events, use that instead of piecing together raw page view reports. If you use measuremy.site, you can build a funnel from page paths and events and see the counts directly. Be aware of how your analytics identifies visitors. Some cookieless tools reset visitor identity every day, which means you can only reliably analyze same-day sequences. If a visitor starts a signup on Monday and finishes on Tuesday, a daily-reset tracker may not connect those steps. Check how your tool handles identity before you trust cross-day funnel numbers. ## Calculate drop-off between every pair of steps For each transition from one step to the next, calculate the [drop-off](https://measuremy.site/blog/conversion-funnel-analysis) as a percentage: Subtract the count at step two from the count at step one, divide by the count at step one, and multiply by 100. Do this for every adjacent pair. Also record the absolute number of people lost. For example, if 1,000 people land on a page and 600 click the button, the drop-off is 40 percent, or 400 people. If 600 click and 500 submit, the drop-off is about 17 percent, or 100 people. ## Separate absolute drop-off from relative drop-off The biggest percentage drop is not always the step that loses the most people. A 50 percent drop from 1,000 to 500 is 500 people. A 70 percent drop from 100 to 30 is only 70 people. Both matter, but they point to different decisions. - The largest absolute drop tells you where the most potential customers are leaving. This is often the first step, simply because the audience is largest there. - The largest relative drop tells you where the experience is most broken for the people who made it that far. - For a small site, prioritize the step that loses the most people, because fixing it can recover the largest number of conversions. ## Check whether the data is complete Before you act on a drop-off number, make sure the step counts are accurate. Common tracking gaps make a real drop look worse than it is. - The analytics tag is missing on one page in the funnel. - A button fires a click event twice, inflating the count before the next step. - A redirect or JavaScript navigation skips a page view. - Bot traffic or test clicks are counted as real visitors. - [UTM parameters](https://measuremy.site/blog/utm-tracking-explained) are lost on a redirect, so the funnel breaks by source. Check your own setup against these points. If a step looks suspiciously low, verify that the tag or event is actually firing on that page. ## Find the biggest real loss Once you trust the numbers, look for the step where the largest number of visitors leave. On many small sites, the biggest drop is between the landing page and the first meaningful action. That is where people decide whether the page matches their intent. But do not ignore a later step with a high percentage drop if enough people reach it. A simple way to compare is to list each transition with its absolute loss and percentage loss, then highlight the largest absolute loss and the largest percentage loss among steps with at least 50 or 100 visitors. That minimum depends on your traffic, but below that the percentages are noisy. ## Look for the cause at that step Once you know which step leaks the most people, investigate what happens there. Check the page in a real browser on mobile and desktop. Look for slow loading, broken buttons, confusing copy, too many form fields, missing trust signals, or unexpected price changes. If you have session recordings or heatmaps, watch a few sessions that ended at that step. Do not assume the cause; use the data to form a hypothesis and test a change. ## Use event funnels to narrow the problem If the drop happens between two pages, add a click event on the main button or link that moves people forward. Then compare how many people see the page with how many click the button. If many see the page but few click, the problem is motivation or clarity. If few see the button at all, the problem may be layout, page speed, or a broken element. This extra step helps you avoid fixing the wrong part of the page. ## Segment by source or device A large overall drop can hide a segment that performs well. If your traffic volume allows, break the funnel down by source, device type, or country. - Compare organic search, social, and email traffic. - Compare mobile and desktop. - Compare the top two or three countries. Do not over-segment. With a small site, splitting a funnel into too many slices leaves each slice too small to trust. Start with one or two segments that you can act on. ## Validate small-sample drop-offs When a funnel step has only a handful of visitors, a big percentage change can come from random noise. A 60 percent drop from 10 to 4 visitors is not the same as a 60 percent drop from 1,000 to 400. If your funnel is small, look at the same drop over several weeks. If the same step consistently loses a similar share of visitors, you have a real pattern. If it jumps around, collect more data before making changes. ## Let a coding agent pull the numbers for you If you use measuremy.site, its [MCP server](https://measuremy.site/guides/analytics-mcp-server) lets a coding agent read the analytics, create funnels and tracked links, and report the step with the largest absolute and relative drop-off. Instead of exporting data and building a spreadsheet, you can ask the agent to calculate the numbers and show the funnel. This is useful when you want to check drop-offs after every change without doing the math by hand. The most reliable way to find the biggest drop off is to define the funnel clearly, collect accurate step counts, calculate both absolute and relative losses, and then investigate the step that loses the most real visitors. Check your tracking before you trust the numbers, and treat small samples as rough signals rather than precise answers. ### See it on your own site measuremy.site is cookieless analytics your coding agent can read. 30-day free trial, no card. Plans from $9/month. [Start free trial](https://measuremy.site/login) [See the live demo](https://measuremy.site/share/measuremy.site) Add measuremy.site analytics to this project: connect the measuremy.site MCP server at https://a.measuremy.site/mcp (it signs in with OAuth, no key), then tell me how to sign in. Details: measuremy.site/llms.txt ## Free tools - [Free conversion rate calculator](https://measuremy.site/tools/conversion-rate-calculator): Conversion rate from visitors and signups, plus the traffic needed for a goal. - [Free A/B test significance calculator](https://measuremy.site/tools/ab-test-significance-calculator): Is your A/B test result real or luck? Check significance. - [UTM link checker](https://measuremy.site/tools/utm-link-checker): Find broken UTM tags and see which GA4 channel a link lands in. [All free tools](https://measuremy.site/tools) ## More from the blog - [How to calculate your website conversion rate, and what to compare it to](https://measuremy.site/blog/calculate-website-conversion-rate) - [Is my A/B test result significant? A plain-English guide with a worked example](https://measuremy.site/blog/is-my-ab-test-significant) - [What is bounce rate, how is it calculated, and is yours a problem?](https://measuremy.site/blog/what-is-bounce-rate) - [How to add a visitor counter to your website or GitHub README](https://measuremy.site/blog/add-visitor-counter-to-website) - [How much traffic does a website get? How to estimate any site's visits](https://measuremy.site/blog/how-much-traffic-does-a-website-get) - [How to track QR code scans: count them, and see which poster worked](https://measuremy.site/blog/track-qr-code-scans) - [How to see how many people clicked your link (without a paid shortener)](https://measuremy.site/blog/how-many-people-clicked-my-link) - [How to check if your Google Analytics 4 tag is installed (and only counting once)](https://measuremy.site/blog/check-if-ga4-is-installed) - [Why GA4 shows your campaign traffic as Unassigned (and how to fix the links)](https://measuremy.site/blog/ga4-unassigned-traffic) - [Does your redirect keep your UTM tags? How to check before a campaign goes out](https://measuremy.site/blog/redirect-drops-utm-tags) - [How I built a self-hosted analytics MCP server on Cloudflare Workers and D1](https://measuremy.site/blog/analytics-mcp-server-cloudflare-workers) - [UTM tracking explained: what the parameters mean and how to use them](https://measuremy.site/blog/utm-tracking-explained) - [Conversion funnel analysis for small sites: a practical guide](https://measuremy.site/blog/conversion-funnel-analysis) - [Sudden drop in website traffic? A checklist for finding the cause](https://measuremy.site/blog/sudden-drop-in-website-traffic) - [What is an MCP server? A plain-English guide for people who run websites](https://measuremy.site/blog/what-is-an-mcp-server) - [What Is the Urchin Tracking Module?](https://measuremy.site/blog/what-is-the-urchin-tracking-module-and-why-utm-stuck-around) - [What Microsoft MCP servers actually do](https://measuremy.site/blog/microsoft-mcp-servers-explained) - [UTM parameters for email newsletters](https://measuremy.site/blog/utm-parameters-for-email-newsletters) - [Track Button Clicks Without Google Tag Manager](https://measuremy.site/blog/how-to-track-clicks-on-a-button-without-google-tag-manager) - [How to track the results of a social media post without cookies](https://measuremy.site/blog/how-to-measure-the-results-of-a-social-media-post) - [How to track traffic from a GitHub README](https://measuremy.site/blog/how-to-track-traffic-from-a-github-readme) - [How to measure a product launch without guessing](https://measuremy.site/blog/how-to-measure-a-product-launch)