A conversion funnel is the map of how a stranger turns into a customer, and most founders are flying without it. It splits the journey into stages, visit, sign up, activate, pay, so you can see the exact step where people quietly vanish. The shape is a funnel because you start wide and lose people at every level. The magic is not the shape. It is that each stage gets its own drop-off number you can attack.
We learned this on our own products the slow way. We poured budget into ads to widen the top, watched signups tick up, and watched revenue stay flat. The leak was not at the top. It was a brutal drop between signup and the first real action, and we could not see it until we drew the funnel and stared at the cliff between two stages.
This guide covers what a conversion funnel actually is, the stages worth using, the real benchmarks per stage, a five-step build, and the free tools that map it. We will also be blunt about what to cut, because the CRO software aisle is full of things you do not need yet.
◢What is a conversion funnel?
A conversion funnel is a staged model of the path from first touch to paying customer, designed so you can measure the drop-off at each step. The concept traces back to the classic purchase funnel of awareness, interest, desire, and action, first sketched out over a century ago. The labels evolved. The core idea did not.
The value is not the pretty pyramid. It is that a funnel forces specificity. "We have a conversion problem" is a shrug. "We lose 60% of people between signup and first project created" is a target with a deadline.
That shift changes what you do on Monday. Instead of widening the top with more spend, you find the worst leak and patch it. A funnel turns a vague worry into a ranked to-do list, which is the whole point.
◢What are the stages of a conversion funnel?
The stages depend on your model, but the rule is constant: each stage is one measurable action with its own conversion rate. Marketers use top, middle, and bottom of funnel. Startups get more mileage from Dave McClure's AARRR pirate metrics: acquisition, activation, retention, referral, and revenue. He built it in 2007, and it still maps cleanly to how products grow.
Acquisition is how people find you, through search, ads, social, or word of mouth. It is the only stage most teams obsess over, which is exactly the trap, as Inc. noted when McClure first made the rounds.
Activation is the first moment a user feels the value, like sending their first message or building their first dashboard. This is where most funnels bleed, and it is the stage worth your attention first. If you are wiring this up, our user onboarding flow recipe maps the moves.
Retention, referral, and revenue are the bottom three: people coming back, bringing friends, and paying. PostHog's breakdown and Amplitude's guide both stress that retention, not acquisition, is where durable growth actually lives. Cut the instinct to dump every dollar into the top.
◢What is a good conversion funnel rate?
There is no universal good rate, because it swings hard by industry, price, and stage. For ecommerce, the global average hovers around 2.5%, per Smart Insights, and top Shopify stores clear 3.2%. For B2B SaaS, end-to-end visitor-to-customer runs roughly 0.5% to 1.5%, with the best teams reaching 2% to 4%, per LeanLabs.
Industry matters more than people admit. Food and beverage stores convert as high as 4.9% to 6.2%, while home decor crawls near 1.4%, per Statsig's breakdown. Comparing your SaaS funnel to a snack brand's is a great way to feel bad for no reason.
So chase your own trend, not someone's benchmark. A funnel that gains a point over last quarter is winning. One that merely matches a blog average is standing still while telling itself it is fine.
◢Where do conversion funnels leak the most?
Most funnels leak hardest in the middle, not the top, which is exactly where teams look least. The single most documented leak in ecommerce is checkout: the average cart abandonment rate is 70.19%, calculated across dozens of studies by the Baymard Institute. Roughly seven of every ten people who add to cart never buy.
Mobile is worse, abandoning near 80% versus 66% on desktop, per Baymard. A clunky mobile checkout is a tax you pay on every single visit. Shopify's own checkout teardown pins the usual culprits: surprise shipping costs, forced account creation, and too many form fields.
For SaaS, the equivalent leak sits between signup and activation. People create an account, hit an empty dashboard, and never come back. We have watched this kill otherwise healthy funnels. The fix is almost never more traffic. It is removing one step of friction at the exact stage where the floor drops out.
◢How do you build a conversion funnel in five steps?
Build a funnel by defining the stages, picking one metric per stage, instrumenting the events, pulling the drop-off, and attacking the worst leak first. The order is load-bearing. Skip straight to tools and you get a dashboard nobody reads.
First, define your stages as real actions in your product, like visit, signup, activate, and pay. Keep it to four or five. A ten-stage funnel is a chart, not a decision.
Second, pick one metric per stage and the conversion rate between each. Third, instrument the events so each stage fires a clean signal. If you are choosing where this data lives, our product analytics stack recipe covers the free options. Fourth, pull the drop-off and rank the stages by how much you lose. Fifth, fix the worst one, ship, and remeasure. Then repeat down the funnel. A small win at a high-traffic stage beats a heroic win at a stage nobody reaches.
◢What conversion funnel tools should you actually pay for?
Honestly, almost none of them at the start. Google Analytics 4 ships a free Funnel Exploration report in its Explore section that shows drop-off at every step, plus path analysis to see where people wander. That covers most early websites for zero dollars.
For product funnels, PostHog's free tier includes 1 million events a month with built-in funnels, paths, and session replay. We have run real funnel analysis on side projects without paying a cent. The free ceiling is high enough that most pre-revenue teams never hit it.
So what do you cut? The bundled CRO suites with per-seat pricing and a sales call before you can even see the product. Heatmaps, A/B testing, and personalization engines are real tools, but they are stage-two purchases. If your funnel has a 70% leak at checkout, you do not need a personalization engine. You need to delete three form fields. Before you commit to any paid analytics or CRO tool, run the numbers through our stack cost calculator, and compare options in our analytics tool alternatives library.
◢The bottom line
A conversion funnel is not a fancy growth concept. It is the cheapest map you will ever own, and you can draw it today with tools you already have. Three things to take with you. First, the leak is almost always in the middle, not the top, so stop reflexively buying more traffic. Second, chase your own trend instead of someone else's benchmark, because a snack brand's 6% means nothing to your SaaS. Third, build it free in GA4 or PostHog before any CRO suite gets your credit card.
We map every funnel this way for our own projects, and the boring version wins every time. Find the worst drop, fix that one stage, remeasure, repeat.
Want the no-kickback teardown of which growth tools are worth paying for? Subscribe to the Cut The SaaS newsletter. Nobody pays us to recommend anything, and we like it that way.
◢FAQ
What is a conversion funnel in simple terms? A conversion funnel is the journey a person takes from first hearing about you to becoming a paying customer, split into clear stages like visit, sign up, and purchase. It is shaped like a funnel because you start with many people at the top and lose some at every step, so fewer reach the bottom. The point of drawing it as stages is to see exactly where people drop off. A flat "we lose customers" is useless. "We lose 70% of carts at checkout" is a fix you can actually ship.
What are the stages of a conversion funnel? The classic marketing model is awareness, interest, desire, and action, often shortened to top, middle, and bottom of funnel. For startups, the more useful version is Dave McClure's AARRR: acquisition, activation, retention, referral, and revenue. The exact labels matter less than the rule behind them: each stage is a measurable step with its own drop-off rate. Pick stages that match a real action in your product, like signup, activation, and paid conversion, so every number maps to something you can change.
What is a good conversion funnel rate? There is no single good number, because it swings wildly by industry and stage. For ecommerce, the global average sits around 2.5%, per Smart Insights, while top Shopify stores clear 3.2%. For B2B SaaS, end-to-end visitor-to-customer runs roughly 0.5% to 1.5%, with the best hitting 2% to 4%. Chase your own trend, not a benchmark. A funnel that converts a point better than it did last quarter beats one that merely matches some blog's average.
Can you build a conversion funnel for free? Yes, and most early teams should. Google Analytics 4 ships a free Funnel Exploration report in its Explore section, per Google's docs, that shows drop-off at every step. PostHog's free tier includes 1 million events a month with built-in funnel and path tools, per its pricing. A spreadsheet works too if your volume is small. We have mapped funnels for side projects without paying anything. Buy a dedicated CRO suite only when the free tools genuinely block you.
What is the difference between a conversion funnel and a sales pipeline? A conversion funnel tracks anonymous and known users moving through stages, often automatically, and is measured in rates and drop-offs. A sales pipeline tracks named deals that a human rep moves through stages by hand, measured in deal value and close dates. The funnel is the top-of-funnel demand machine. The pipeline is what sales does with the qualified leads it spits out. Product-led companies live in the funnel. Sales-led ones lean on the pipeline. Most growing startups end up running both.