Funnel analysis is the practice of mapping each step a visitor takes toward a goal — signup, purchase, upgrade — and measuring how many people complete each step versus how many drop off. Instead of looking at a single conversion rate as one number, you break the journey into stages and find exactly where you're losing people.
Most websites lose the majority of visitors before they convert. The question isn't whether you have drop-offs — you do. The question is which drop-off is costing you the most, and whether you can fix it.
What a Conversion Funnel Looks Like
A funnel is a sequence of actions that lead to a goal. For a SaaS product, a typical funnel might be:
- Visit landing page — the visitor arrives
- View pricing — they're evaluating
- Start signup — they've decided to try
- Complete signup — they finish the form
- Activate — they take the first meaningful action in the product
Each step filters out a portion of visitors. By measuring the conversion rate between each pair of steps, you find the biggest leak.
Here's what a real funnel might look like:
| Funnel Stage | Visitors | Drop-Off | Stage Conversion |
|---|---|---|---|
| Landing page | 10,000 | — | — |
| View pricing | 3,200 | 68% left | 32% |
| Start signup | 960 | 70% left | 30% |
| Complete signup | 720 | 25% left | 75% |
| Activate | 288 | 60% left | 40% |
Overall conversion from landing page to activation: 2.9%. But the biggest leaks are at the top — 68% of visitors never look at pricing, and 70% of those who do never start signing up.
Key Summary
The largest absolute drop-offs usually happen early in the funnel, where visitor volume is highest. Fixing a 5% improvement at the top of a 10,000-visitor funnel adds 500 people to the next step. The same improvement at the bottom adds far fewer.
How to Build a Funnel
The mechanics depend on your analytics tool, but the process is the same.
1. Define the goal. What counts as a conversion? For SaaS, it's usually a completed signup, a started trial, or a first purchase. Be specific — "signed up" and "activated" are different goals with different funnels.
2. Map the steps backward. Start from the conversion event and work backward through the pages and actions that precede it. Don't guess — look at your actual data. What pages do converters visit before converting? That sequence is your funnel.
3. Set up event tracking. Each funnel step needs a trackable event: a page visit, a button click, or a form submission. If you're using Trfyx, pageview-based funnels work out of the box. For click-based steps, add custom events.
4. Set a time window. Decide how long a visitor has to complete the funnel. Same session? Seven days? Thirty days? A shorter window is stricter but gives you faster feedback. Most SaaS funnels use a 7- to 14-day window.
Finding Drop-Off Patterns
Once your funnel is running, the data shows you where to focus. But the numbers alone don't tell you why people leave. You need to combine funnel data with other signals.
High drop-off on the pricing page. If most visitors leave after seeing your pricing, the problem could be price shock, confusing plan tiers, or a missing plan that matches their needs. Check whether visitors scroll to the bottom of the pricing page — if they don't, the layout might be the issue, not the price.
Drop-off between signup start and completion. If visitors begin the signup form but don't finish, watch session replays of those sessions. Common culprits: too many form fields, a required phone number, a broken field on mobile, or a confusing password requirement.
Low activation after signup. The visitor signed up but never used the product. This is an onboarding problem, not a website problem. But you can still use analytics to find the gap — what's the first action active users take? If 80% of retained users complete a setup wizard within the first session but only 40% of signups reach it, the wizard might be hard to find.
Device-specific drop-offs. Compare your funnel on desktop versus mobile. If a step converts at 45% on desktop but 15% on mobile, you likely have a layout or usability issue on small screens. Check heatmaps for that page on mobile to see what visitors are interacting with.
Fixing Each Funnel Stage
Different stages call for different fixes.
Top of funnel (landing → pricing/features). The visitor arrived but isn't interested enough to explore. Possible fixes:
- Improve headline clarity — does the visitor immediately understand what you do?
- Add social proof above the fold
- Make the CTA more visible
- Reduce page load time — every second of delay reduces conversions
Middle of funnel (pricing → signup start). The visitor is evaluating but hasn't committed. Possible fixes:
- Simplify the pricing page — fewer plans, clearer feature comparisons
- Add a free tier or trial to lower the commitment threshold
- Include testimonials from users in the visitor's segment
- Show what happens after signup (a preview of the dashboard or onboarding flow)
Bottom of funnel (signup → activation). The visitor is sold but struggling to complete the process. Possible fixes:
- Reduce signup form fields to the absolute minimum
- Remove friction (email verification can happen after first use, not before)
- Send a targeted onboarding email within minutes of signup
- Add a progress indicator to multi-step flows
Common Funnel Analysis Mistakes
Too many steps. A funnel with 12 steps is hard to analyze and harder to fix. Focus on 4 to 6 key steps that represent real decision points, not every page in the flow.
Ignoring micro-conversions. The main funnel matters most, but secondary actions — watching a demo video, downloading a guide, clicking "see all features" — indicate intent. Track these as separate funnels or as supporting metrics.
Only looking at percentages. A 50% drop-off at a step with 100 visitors costs you 50 people. A 10% drop-off at a step with 10,000 visitors costs you 1,000. Always consider absolute numbers alongside percentages.
Not segmenting. Your overall funnel might show 3% conversion, but visitors from organic search might convert at 6% while paid social converts at 0.5%. Segment by traffic source, device, and geography to find the real story. This is where connecting funnel data to revenue attribution becomes valuable — not just who converts, but who converts and pays.
Getting Started
If you don't have a funnel set up, start with one. Pick your most important conversion goal, identify the three to five steps that lead to it, and set up tracking. Even a basic funnel will show you something you didn't know.
Trfyx includes funnel visualization as a core feature — define steps using page URLs or custom events, set a time window, and see the drop-off at each stage. From there, click into any stage to see the sessions that dropped off, or jump to session replays to watch what happened.
The first funnel you build won't be perfect. You'll realize you're missing a step, or that a step you included isn't meaningful. That's fine — refine it based on what the data shows. The goal is to stop guessing about where you're losing customers and start measuring it.
Summarize with AI
Get a quick summary of this article using your favorite AI assistant.