What Is Funnel Analysis?
Funnel analysis is one of the most practical tools in digital analytics. It models a specific user journey, typically a sequence of pages or events, and shows you what percentage of users make it from one step to the next. The visual output looks like a funnel because, as users progress, the number who continue narrows at each stage. The critical question funnel analysis answers is: where are we losing people, and why?
A standard e-commerce purchase funnel might include five steps: product page view, add to cart, checkout initiation, payment details entry, and order confirmation. If 1,000 users view a product page but only 80 complete a purchase, the overall conversion is 8%. Funnel analysis breaks that journey apart, showing that 700 users add to cart (70%), 300 initiate checkout (43% drop from cart), 120 enter payment details (60% drop from checkout), and 80 complete the order. The biggest problem is clearly the checkout-to-payment step, not the product page.
Funnel analysis is available in Google Analytics 4 using the Funnel Exploration report, and in dedicated product analytics platforms like Mixpanel, Amplitude, and Heap. In GA4, funnels can be open or closed. Open funnels count users who enter at any step, giving a broader view of natural user paths. Closed funnels only count users who start at step one, which is more useful when evaluating a specific designed journey like an onboarding flow.
For South African businesses, funnel analysis is particularly important for checkout flows, because local payment friction points differ from global norms. Mobile payment preferences, load shedding interruptions mid-session, and distrust of unfamiliar payment gateways create local drop-off patterns that only granular funnel data can reveal.
Funnel Analysis In Practice
A Durban-based online fashion retailer runs funnel analysis on their checkout and discovers that 62% of users who add an item to their cart abandon at the shipping cost reveal step. The team tests a "free shipping over R750" threshold and finds that cart abandonment on this step drops to 38%. The revenue increase from the extra orders more than offsets the free shipping cost. Without funnel analysis, the team would have guessed incorrectly that the problem was their payment gateway rather than price transparency at the shipping step.
Funnel analysis also reveals segment differences that aggregate web analytics data hides. A Johannesburg B2B software company discovers that mobile users drop off at the trial sign-up form at a rate 3x higher than desktop users. Inspecting the form on mobile, the team realises one field requires file upload, which is nearly impossible on a phone. Removing that field from the mobile experience increases mobile trial sign-ups by 40% in the first month. This kind of targeted, step-specific insight is exactly what makes funnel analysis so valuable alongside tools like heatmaps and session recordings.
How to do funnel analysis
Funnel analysis measures how many people progress through each defined step towards a goal, and where they drop out. You start by mapping the steps a visitor should take, say landing page, product page, add to basket, checkout, purchase, then measure the conversion rate between each pair. The value is in the gaps: a sharp fall between two steps localises the problem to that transition, whether an unclear next step, an unexpected cost, or a slow page. Rather than judging only the overall conversion rate, funnel analysis shows exactly which step leaks most, so you fix the specific bottleneck that costs the most conversions instead of optimising blindly across the whole journey.
Reading a funnel and acting on it
A funnel's numbers only help when they change what you do. Once you see the step with the biggest drop, investigate why: watch session recordings, check the page on mobile, look for friction such as forced account creation or costs revealed late. Then fix that step and re-measure, so you know whether the change worked. It also helps to segment the funnel, by device, source or audience, since a leak may affect only mobile users or one traffic source, which averages hide. Analytics tools model funnels directly, but the discipline matters more than the tool: find the worst leak, understand it, fix it, and confirm the fix, repeatedly.
FAQ
What is a good funnel conversion rate for South African e-commerce?
Average e-commerce conversion rates in South Africa range from 1% to 3%, depending on product category and traffic source. Organic and email traffic typically converts at the higher end, while cold social media traffic often falls below 1%. Funnel analysis helps identify which step is limiting your overall rate so you can focus improvements where they matter most.
How is funnel analysis different from cohort analysis?
Funnel analysis measures the percentage of users completing each step in a defined sequence, highlighting drop-off points at a single point in time. Cohort analysis follows a group of users across multiple time periods to track how behaviour changes after a starting event. Both are complementary tools used together in conversion rate optimisation programmes.
What does funnel analysis show that overall conversion rate does not?
It shows where in the journey people drop out, not just how many convert overall. A single conversion rate hides which step causes the loss; funnel analysis localises it, so you can fix the specific transition, checkout, say, that leaks most rather than guessing across the whole path.
What tools are used for funnel analysis?
Analytics platforms such as Google Analytics 4 can model funnels, and behaviour tools like heatmaps and session recordings help explain why a step leaks. The tool matters less than defining the right steps and acting on the biggest drop-off you find.