A conversion rate is meaningful only when its denominator matches the question. State the eligible population, counted event and observation window, retain both counts and distinguish people from sessions or opportunities. Compare rates only after checking that exclusions and maturity are consistent.
Write the question as a fraction
Before opening a dashboard, write a sentence such as “Of eligible enquiries received in June, how many accepted a proposal by the review cutoff?” The denominator is those enquiries; the numerator is the subset meeting the stated event. This simple step prevents a common error: dividing purchases from several months of sales work by this month’s visits and presenting the result as one coherent journey.
Choose the unit deliberately. Sessions describe visits, users describe identified people and organisations may contain several contacts. Opportunities can include repeat purchases from one organisation. None is inherently the correct unit for every question. Keep names precise and avoid switching units midway through a report. If identity is incomplete, use a label such as eligible sessions and explain the limit rather than promoting an estimate into an exact count of people.
Check eligibility, timing and duplicate records
Define exclusions that follow from the question, including staff tests, known automated traffic or requests for an unrelated offer. Apply them consistently before comparing periods. Keep an exceptions record when eligibility is uncertain. Do not remove unsuccessful cases after seeing their outcome merely because they make the rate lower. A useful denominator represents the opportunities that actually faced the transition you are evaluating.
Align observation windows. A proposal issued yesterday has had less time to become a purchase than one issued two months ago. Use a fixed cutoff with maturity notes or compare cohorts at the same age. Deduplicate repeated form submissions using an appropriate stable identifier and retain the original event history. If one enquiry creates several proposals, decide whether the measure follows enquiries or proposals and document that choice beside the calculation.
Display counts and limitations with the percentage
Show numerator and denominator next to the rate. “Three of six” communicates uncertainty that “50%” alone conceals. For small counts, discuss actual cases before drawing a trend. Distinguish a percentage-point change from a relative percentage change, and avoid claiming significance unless the analysis supports it. Correct data errors transparently rather than silently replacing a historical number in every slide.
When comparing channels, check that each channel’s population has a similar opportunity to complete the action. A referral conversation and an anonymous page visit are different starting points. A high late-stage conversion rate may simply reflect prior qualification. SBA marketing guidance encourages consistent measurement across tactics; matching the denominator is the practical discipline that makes such reviews interpretable. Record which business decision the rate will inform.
Keep an owner and calculation note for each important metric. If a colleague cannot reconstruct it from the stated records, the definition needs work before automation. Use a small manual sample after any tracking change. A technically valid formula can still answer the wrong question if a form event is now recorded twice or a new route bypasses the original eligibility check.
Fictional worked example
A fictional scenario to explain the method, not a customer result or a market benchmark.
In a fictional advice service, 200 eligible landing-page sessions produce twenty enquiry submissions. Two are duplicate submissions from the same opportunities, leaving eighteen distinct enquiries. The team reports a 10% session-to-submission rate and separately an 18-enquiry count. It does not call eighteen divided by 200 a person-level rate because the session data does not identify unique people reliably.
By the review date, six of those eighteen enquiries have purchased, so the observed enquiry-to-purchase rate is one third, approximately 33.3%. Four enquiries remain open and are labelled accordingly. The team retains the cutoff rather than presenting the open cases as settled losses. These invented figures demonstrate why the unit and maturity matter; they are not a recommended conversion target or a commercial forecast.
TAKE THIS INTO YOUR NEXT CONVERSATION
A practical next-step checklist.
- Write the rate as a question and fraction.
- Name the unit, eligibility rule and observation cutoff.
- Deduplicate without removing legitimate unsuccessful cases.
- Display counts and preserve a reproducible calculation note.
Common questions.
Can the same report contain session and customer rates?
Yes, provided each is clearly labelled with its own denominator. Do not imply that separate aggregates represent a fully traced person-level journey.
Should open opportunities be excluded?
Choose a rule that matches the question. Usually retain them in an entry cohort and show their unresolved status or compare only equally mature cohorts.
Further reading & source context.
These primary references provide background. They are not evidence for the fictional example above and do not endorse AccelerateLink.
Found an error or a missing qualification? Send an editorial correction.
Find my next move.
Use a practical tool to turn this idea into a clearer starting point.
Find my next move