The Answers Are Already in the Room

Bookings, cancellations, payments and service durations can help answer a practical operating question. Start with that question rather than buying another reporting tool or asking the team to collect everything.

Existing records may be enough, but missing history, inconsistent setup and access restrictions can limit the answer. Establish what the source systems can reliably show.

Start with one operating question

Choose a question tied to a decision: which appointment gaps are genuinely fillable, which new-client groups return within an eligible window, or whether a service covers its direct costs.

• Acquisition payback: use collected revenue, direct costs and known acquisition cost, not visit count alone.
• Service pathways: compare observed repeat behavior while allowing for service cadence and an incomplete observation window.
• Capacity: distinguish staffed, usable time from breaks, preparation and gaps too short to sell.

Identify the records and definitions needed

For a capacity question, the minimum useful inputs might be scheduled practitioner hours, actual service durations, planned buffers, cancellations and suitable demand. Confirm access, time zones, record completeness and consistent service identifiers before combining sources.

Deliver a plain-language finding with the calculation and its limits. Offer a portal or deeper record review when useful; a concise summary should not hide evidence someone needs to inspect.

Use records only for the agreed purpose

Map what is accessed, which necessary providers process it, who can review it and how long it is retained. An API connection does not mean that no operational records leave the source system or that nothing is stored.

The Privacy Policy explains the current purpose, provider, non-sale and retention boundaries. The project agreement and actual deployment settings determine the specific permitted fields and actions. Do not extend an ownership slogan into a promise that every system supports the same export or deletion behavior.

Test what the records can establish

Ask for a small review with a defined decision, source list and sample calculation. Mark missing values as unknown. Compare the finding with staff knowledge before acting, then measure whether a bounded process change improves the relevant outcome.