Skip to main content

What Is a Sales Funnel, and Where Does the Model Break Down?

Summarize with AI

A four stage sales funnel showing suspects narrowing to prospects, leads and customers

Knowing where a model fails is what makes it safe to use, so the second half of this article is about exactly that.

The four stages

StageWho they areWhat moves them forward
SuspectsEveryone who might plausibly want what you sellBeing visible where they already are
ProspectsPeople showing any signal of interestSomething worth their attention
LeadsPeople who have identified themselves as interestedA reason to choose you specifically
CustomersPeople who paidDelivery, and a reason to come back
The four classic funnel stages and the thing that moves people between them

The coffee shop version makes it concrete.

Everyone walking past your door is a suspect. You do not know whether they want coffee. The ones who stop to read the menu or pull into the lot are prospects: they have shown a signal. The ones who walk in are leads. The ones who order and pay are customers.

Online it works the same way with different signals. Everyone searching for what you sell is a suspect. Everyone who lands on your site is a prospect. Everyone who fills in a form, calls, or starts a chat is a lead. Everyone who buys is a customer.

The value of the model is that it forces one useful question: at which stage are people falling out? A business losing people between prospect and lead has a website problem. A business losing them between lead and customer has a sales problem. Those need completely different fixes, and without a funnel to hang the numbers on, both look identical from the outside.

That question alone justifies the model.

Four places the funnel breaks

Now the part that vendor content skips, because most companies explaining funnels are selling funnel software.

1. People do not move in one direction

The funnel is drawn as a one-way trip. Real buying loops. Somebody discovers you, disappears for three weeks, comes back and compares you against two competitors, leaves again, asks a friend, sees a retargeting ad, then converts.

Google's own research on consumer decision making describes this middle section as a loop between exploring and evaluating, repeated as many times as the buyer needs, rather than a sequence of stages. Anybody who has watched a real B2B deal or a $12,000 home services quote knows this is accurate.

How real buying journeys loop between exploration and evaluation instead of moving down a funnel in one direction
The loop that the funnel diagram cannot draw

The practical damage: if you treat stages as one-way, you build sequences that assume forward motion and stop talking to people who stepped back. Those people were not lost. They were shopping.

2. Most of the journey is invisible to you

The funnel implies you can see people move through it. You cannot see most of it.

Somebody hears about you in a WhatsApp group, checks your Instagram, reads a review on a site you have never heard of, asks a neighbor, then types your business name into Google two weeks later. Your analytics records one visit: branded search, direct conversion.

Every touchpoint that actually did the work is invisible, and your reporting confidently credits the last one. This is not a tracking failure you can fix with better tags. Most of it happens in places that do not report to anyone.

The damage: channels that build demand look worthless in the report, and the channel that captures already-existing demand looks like a genius. Cut the first to fund the second and you get a good quarter followed by a bad year.

We wrote about the related tracking problem in how to track website traffic across channels, and the honest conclusion there is the same one: your data is a partial view being presented as a complete one. The part you can control is tagging your own links, which is what UTM parameters and why attribution lies to you covers.

3. The funnel treats a customer as the end

The narrow end of the funnel is the purchase, which is exactly backwards for most businesses. Retention, repeat purchase and referral are where the margin lives, and the shape of the diagram puts them off the page.

For a local service business this is severe. A satisfied customer who tells three neighbors is worth more than any single channel in your acquisition mix, and the funnel model has nowhere to draw them.

4. It flattens very different journeys into one shape

The four stages describe an emergency plumber call and a six-month software evaluation with the same picture. One of those is a five minute decision with one decision maker. The other has a committee, a budget cycle and a procurement process.

Using one funnel for both means your stage definitions are too vague to act on for either.

So what do you do with it

Keep the funnel. Use it for the one thing it does well and stop asking it to do more.

Use it as a diagnostic. Count how many people are at each stage this month and compare to last month. Where the drop-off changed is where to look. That is the entire job and it is genuinely valuable. The conversion path reports in GA4 give you part of the picture, and what GA4 reports show about conversion paths explains which ones.

Stop using it as a description of behavior. People will not move through it in order, and building automation that assumes they will produces emails that arrive at the wrong moment and make you look like you are not paying attention.

Add a loop after the sale. Whatever you do to acquire a customer, budget something to keep them and something to make referral easy. If it is not on the diagram it will not get funded.

Define the stages for your actual business. "Prospect" has to mean something you can count. For a local service business it might be "requested a quote". For ecommerce it might be "added to cart". Vague stages produce vague numbers.

Which channel serves which stage is its own decision, and we worked through it with real budgets in paid media strategy for Miami businesses.

Common questions

What are the stages of a sales funnel?

The classic model has four: suspects, prospects, leads and customers. Many versions use awareness, interest, decision and action instead, which is the same idea with different labels. The number of stages matters far less than whether each one is defined so you can count the people in it.

What is the difference between a marketing funnel and a sales funnel?

A marketing funnel usually covers the top, from first exposure to a qualified lead. A sales funnel usually covers the bottom, from qualified lead to closed deal. In small businesses the same person owns both, which is why the terms get used interchangeably.

Is the sales funnel still relevant?

As a diagnostic, yes. As a description of how people actually buy, it has been outdated for years, which is why the linear model gets criticized so heavily. Use it to find where you are losing people, not to predict what any individual will do next.

How many touchpoints does it take before someone buys?

The commonly quoted figures range from five to twenty depending on price and industry, and all of them are unreliable, because most touchpoints are never recorded. The number is higher than your analytics shows. That is the only part you can say with confidence.

Do I need funnel software?

Not to start. A spreadsheet with four numbers, updated monthly, gives you the diagnostic value. Software becomes worth it when the volume of leads makes manual follow-up unreliable, which is a different problem from understanding your funnel, and it is where our marketing automation service picks up.

The version worth keeping

The funnel is a counting tool wearing the costume of a behavioral theory. As a counting tool it is excellent, and every business should have those four numbers for last month.

As a theory of how people decide, it describes a buyer who does not exist: one who finds you once, moves forward in a straight line, buys, and stops mattering.

Hold both of those at the same time and the model earns its place. Believe the diagram literally and you will build a marketing system for an imaginary customer.

If you want to know where your own drop-off actually is, the exercise takes an afternoon: four numbers, defined precisely, compared to the month before. We do this at the start of most engagements because it usually reveals that the problem everybody assumed was a traffic problem is a follow-up problem.

Find your real drop-off

Four numbers, defined precisely, compared to last month. We run the exercise with you in an afternoon.

Written by J Raydel Sanchez, Founder and CEO of tamer, a digital marketing agency in Hialeah, Florida.

Read Next