A trial tracker should distinguish requested, issued, started, completed and converted-to-paid states. These are different events, and a message asking for a trial is not a sale. Keep actual observations and dates rather than reporting every enquiry as a customer or claiming a conversion rate without a defined denominator.
Define the stages before counting
A trial can be requested but never activated, or issued but never used. Decide what each status means in your records. Use the same definition across reporting periods so the figures remain comparable.
Collect only useful contact data
A nonsecret customer reference and relevant setup details can be enough for tracking. Do not store passwords or unnecessary device identifiers in the marketing report. Keep the actual service access record separate.
Report uncertainty honestly
A customer who stops replying has an unknown outcome, not an automatic failure or success. If you calculate conversion, state the time window and cohort. Avoid using small or incomplete samples as proof of guaranteed future performance.
A practical sequence
- Define each trial stage.
- Record actual status changes and dates.
- Calculate outcomes only from a clearly defined cohort.
An example
Ten trial requests, eight issued trials and three paid orders are three different counts. Calling all ten requests “new paying customers” would misrepresent the record.
Questions about this topic
Is a trial request a paid conversion?
No. It is an earlier stage.
Can silent customers be counted as successful conversions?
No. Their outcome is unknown without evidence.
