"Funnel" has been used to sell so much software that the underlying idea has become obscured. It is not a product, a tactic, or a sequence of automated emails.
A funnel is a description of a simple fact: more people encounter your business than buy from it, and they fall away at identifiable points. Drawing that shape tells you where the largest loss is, which tells you what to work on next.
That is the entire value. Everything else sold under the name is optional.
The four stages, plainly
They do not know you exist. Most of your potential market, at any time.
They know you exist but have not engaged. They have seen you, visited once, followed. Nothing has happened.
They have engaged. Enquired, asked a price, booked a call. They are considering.
They have bought.
Different businesses need more granularity — a long B2B sale has stages inside "considering". Most do not. Four is enough to find the problem.
The only question worth asking
Where does the biggest drop happen?
That is what a funnel is for. Not to build a diagram, but to locate the constraint.
Three businesses with the same revenue can have completely different constraints:
- One is invisible. Everyone who finds them enquires; almost nobody finds them. → discoverability
- One has traffic and no enquiries. → conversion
- One has enquiries and does not close them. → sales
The same investment produces completely different results in each. Spending on advertising when the constraint is conversion buys more people to lose, which is the argument in what to fix before you spend on ads.
Most businesses guess their constraint. It is usually knowable.
Building the picture with numbers you already have
You do not need software for a first version. For last month:
- Roughly how many people encountered you — sessions, reach, footfall
- How many took a first action — enquiry, call, message
- How many became customers
- What a customer was worth
Two ratios come out of that: encounter to enquiry, and enquiry to customer.
Those two numbers locate the constraint immediately. A low first ratio with a healthy second means the site or offer is losing people who were interested. A healthy first with a low second means the problem is qualification or follow-up, not marketing.
That is a more useful hour of work than most funnel software provides.
Fix in the order of leverage
Not every stage deserves equal attention. Improvements compound differently.
Conversion first, usually. It multiplies every channel simultaneously, it is typically the cheapest to change, and the improvement persists.
Follow-up second. If enquiries are not answered quickly, everything upstream is being wasted. This is often the largest single loss and is rarely thought of as marketing.
Discoverability third. Slow, compounding, worth starting early precisely because it takes months.
Paid last. It amplifies whatever exists, including the faults.
The common sequence is exactly the reverse, because paid produces a number fastest.
What funnels are not
Not a set of automated emails. Automation can support a funnel. It is not the funnel, and buying it does not create one.
Not always linear. People move backwards, disappear for months, arrive already convinced by a referral. The model is a simplification, useful for finding gaps, not a description of any individual's path.
Not the same as attribution. The funnel tells you where people fall away; attribution tells you where they came from. Both useful, different questions. Which marketing numbers are worth tracking covers the measurement side.
Not a reason to add stages. Every additional step is another place to lose people. The best funnels are short.
The stage most businesses ignore
After purchase.
Existing customers are the cheapest source of additional revenue and referrals, and for most businesses they receive the least deliberate attention — because the funnel diagram ends at the sale.
Repeat purchase, referral, and simply staying in contact are usually higher-return than acquiring a stranger. They rarely get a budget line because they do not feel like marketing.
If your funnel work produces one change, making it something after the sale is frequently the highest-return option available.
How Truffaire uses this
Our position is instrument before optimising, and the funnel is the simplest possible instrument: it turns "marketing isn't working" into a specific, locatable claim.
In practice, the first thing we establish in a growth engagement is those two ratios. They determine whether the constraint is discoverability, conversion or follow-up — and therefore whether any spend on channels is warranted at all. Recommending a channel before knowing that is a preference presented as expertise.
It is also frequently how we tell a business that the answer is not to hire us for advertising.
Frequently asked questions
Do we need funnel software?
Not to start. Two ratios calculated from numbers you already have will locate your constraint. Software helps at scale and when tracking individuals through longer cycles.
What is a good conversion rate between stages?
Varies too much by sector and traffic source for a benchmark to be meaningful. Your own trend over time is the useful comparison.
Our sale takes months. Does this still apply?
Yes, with more stages between enquiry and purchase, and longer lags. The principle is unchanged: find where people fall away.
Should we have different funnels for different services?
If the buying process genuinely differs, yes. A low-value repeat purchase and a considered project sale behave differently and averaging them hides both.
What if we get most business by referral?
Then your funnel is short and your constraint is likely referral volume rather than awareness. That changes the priority entirely — and it is the kind of thing this exercise reveals.
Where to start
Take last month. Write down four numbers: how many encountered you, how many enquired, how many bought, and what a customer is worth.
Calculate the two ratios. Whichever is weaker is your constraint, and it is probably not the thing you were about to spend money on.
That is a funnel. Everything beyond it is refinement.
Distribution is a system, not a budget covers the layers this maps onto. If you want help locating your constraint, talk to us.