What is a sales funnel?
A sales funnel is a model of how prospective customers move from an initial stage toward buying. Its stages should describe meaningful progress for the buyer, not simply activities the seller completed. Sending a message is different from a person showing interest, and opening a trial account is different from receiving a useful result.
Salesforce’s overview describes a broad progression from awareness through evaluation to purchase while acknowledging that stages vary. A small SaaS does not need to copy an enterprise sales process. Define the decisions and observable actions that fit your actual buying path, then make those definitions usable by whoever records the data.
For a self-service product, that path may include a qualified signup, first task, and subscription. For a founder-assisted sale, it may include a fit conversation, an agreed pilot, an evaluation, and payment. Keep separate paths when combining them would create a stage that means different things for different people.
Sales funnel vs sales pipeline
A cohort funnel can ask what proportion of a starting group reached each later stage within a defined period. A pipeline snapshot can show which opportunities are currently being worked and what happens next. Both are useful, but a list of current open deals is not the denominator for conversion among everyone who entered last month.
Suppose three of twenty interested accounts paid within their completed observation windows. That is a 15% interested-to-paid conversion for that cohort and rule. It does not mean the three remaining deals visible in a pipeline have a 15% probability of closing, nor does it tell you how much revenue next month will bring.
Salesforce’s pipeline guidance organizes work around qualification and the next decision in a deal. Use that operational view to assign follow-up. Use the funnel view to examine progression for a clearly defined population, keeping lost, inactive, and unfinished cases visible.
Define five SaaS funnel stages with entry criteria
Our original fictional BriefBoard model starts with twenty accounts that discussed a relevant brief-preparation problem. It extends the invented pilot in the go-to-market guide by adding an explicit qualification stage. It contains no ShowMeYourSaaS customer or sales results.
The unit is one account, not messages, people, or sessions. Each account is counted once at each stage it reaches in the stated order. The synthetic path requires a trial and first useful result before payment; that is a teaching assumption, not a rule that all SaaS buyers follow. A real product may need additional or optional paths.
Every account in the workbook has a complete 28-day observation window measured from its interested-stage entry. The starting accounts are not a website-visitor population, so the table cannot establish awareness-to-sale conversion. Write those boundaries above the report.
| Stage | Entry criterion in this example | Accounts reaching it |
|---|---|---|
| Interested | Account describes the relevant recurring brief-preparation problem | 20 |
| Qualified | Job, owner, and a feasible near-term evaluation are confirmed | 12 |
| Trial | Account begins the agreed trial workflow | 8 |
| First useful result | Account completes and can retrieve one usable project brief | 5 |
| Paid | First workspace payment is recorded within the observation window | 3 |
Calculate stage conversion and total conversion
Stage-to-stage conversion divides accounts reaching the next stage by accounts reaching the preceding stage, using the same eligible population and conversion window. Interested-to-qualified conversion is 12 ÷ 20 = 60%. Qualified-to-trial is 8 ÷ 12 = 66.67%, rounded. Keep the counts beside the rounded percentage.
Trial-to-first-result conversion is 5 ÷ 8 = 62.5%, and first-result-to-paid is 3 ÷ 5 = 60%. The overall interested-to-paid rate is 3 ÷ 20 = 15%. Trial-to-paid is a different valid view: 3 ÷ 8 = 37.5%. The difference between these percentages is the starting population, not a contradiction.
For a nested sequential funnel, multiplying exact stage fractions gives the overall fraction. Rounded displayed percentages may differ slightly when multiplied. If your data includes people skipping stages or entering halfway through, this simple multiplication no longer represents the same path without additional rules.
| Conversion | Calculation | Result |
|---|---|---|
| Interested to qualified | 12 / 20 | 60% |
| Qualified to trial | 8 / 12 | 66.67% |
| Trial to first useful result | 5 / 8 | 62.5% |
| First useful result to paid | 3 / 5 | 60% |
| Interested to paid | 3 / 20 | 15% |
| Trial to paid | 3 / 8 | 37.5% |
Check identity, ordering, and unfinished windows
Three contacts from one company should not become three accounts merely because each replied to a message. Conversely, one person operating two independent buying accounts may require two records. Choose the unit that matches the sale and define how duplicates, merges, and test accounts are handled.
A stage event count is not automatically a distinct-account count. A person can open a page repeatedly, a webhook can retry, and an account can restart a trial. Salesforce’s funnel-report documentation explicitly distinguishes progression of individuals from the number of times an action occurred. Keep the equivalent distinction in your own instrumentation.
Recent entrants need time to reach the outcome. If a trial started yesterday and the conversion window is 28 days, it is still incomplete for a completed-window rate. Either compare mature cohorts at the same age or show a clearly labeled in-progress view. Do not quietly classify every unfinished account as a failed conversion.
Investigate the next constraint instead of maximizing every percentage
A drop-off identifies a place to investigate. It does not explain the reason. Accounts may be a poor fit, have a later evaluation date, encounter a technical problem, or complete the task elsewhere. Review a small set of representative cases and separate the account’s stated reason from your interpretation.
The synthetic funnel has three fewer first-result accounts than trial accounts. A useful next question is whether those three could provide the required inputs and recognize successful completion. That may lead to an onboarding or product investigation rather than more acquisition. The activation guide helps define the actual first useful result.
A higher percentage can also come from rejecting more prospects or narrowing the denominator. Review qualified demand, delivery effort, and customer outcomes together. Use the CAC guide for the cost of acquiring customers; a conversion-rate table alone cannot establish channel profitability.
Download the sales funnel workbook and template
The original JSON workbook contains all twenty synthetic accounts, their deepest stage, complete-window flags, entry rules, and expected counts. The editable template covers stage definitions, identity checks, conversions, and a next investigation. Sources were checked on October 27, 2026.