TL;DR
Maintain separate event streams for examination relationships and optical transactions. A retail order, collection or adjustment is not evidence of a new clinical relationship. When records can appropriately connect the streams, report the linked cohort and its limits. When they cannot, keep the totals separate rather than manufacturing a conversion funnel.
Give the report separate business units
An optometry business can sell eyewear to someone whose examination happened elsewhere. A patient can attend an examination and buy glasses elsewhere. An existing patient can purchase another product without starting a new clinical relationship. These conditional possibilities are enough to show why a combined 'new customers and patients' total needs explanation. They are not claims about how often those journeys happen.
Choose the reporting question before reconciling records. Are you evaluating newly attended examination relationships, optical transaction activity, or optical orders associated with a defined examination cohort? Each has a different unit and evidence requirement.
For applicable U.S. prescribers, the FTC Eyeglass Rule prohibits requiring agreement to purchase ophthalmic goods as a condition of a refractive examination. The separation provision reinforces why an exam and a retail purchase should not be treated as an inseparable commercial event. It does not prescribe your marketing dictionary.
Adopt an examination and optical event dictionary
| Stream | Event label | Evidence owner | Meaning and exclusion |
|---|---|---|---|
| Examination | Inquiry received | Scheduling | Request arrived; not confirmation or attendance |
| Examination | Appointment confirmed | Scheduling | Relevant exam appointment exists; reschedules remain in its history |
| Examination | Examination attended | Authorized practice reporting owner | Attendance verified for the declared type; prior relationship still needs classification |
| Examination | New clinical relationship | Practice reporting owner | Attended event meets approved historical relationship rule; unknown history stays unknown |
| Optical | Product inquiry | Optical team | A shopping question; no examination outcome established |
| Optical | Order placed | Optical and finance | Order recorded; cancellation, refund and delivery remain distinct |
| Optical | Product collected or delivered | Optical team | Fulfillment event; not another order or acquired person |
| Optical | Adjustment, repair or order-status contact | Optical service team | Service workload; does not create acquisition credit |
Define separate prior-relationship rules for the clinical and optical streams and anchor each to the original inquiry or other declared entry event. Preserve that starting status as later attendance or purchases occur. Otherwise, a report can relabel the same person merely because an order was completed before the next review.
Add the actual system field, reporting period, duplicate rule and unavailable-data treatment. Treat this as a proposed management dictionary, not billing terminology or a software capability claim. Finance should determine which transaction and revenue measures are appropriate.
Keep product units separate from people. Frames, lenses and repeat orders can appear as line items without representing additional acquired patients. An order later remade or refunded needs its own status rather than disappearing from the audit or becoming fresh acquisition. Do not interpret a retail system's 'new customer' label as proof that the clinical records show no prior relationship.
Use a join only for a question it can answer
A proposed exam-to-optical analysis needs a declared examination cohort, a defined later order event, an observation cutoff and an appropriate internal matching method. Have authorized staff do that work inside approved systems. Do not send patient-level clinical or purchasing records to an agency just to fill a chart. The practice's privacy owner should assess the actual summary and recipient before sharing it.
Record the connection as confirmed, unresolved or unavailable. A shared household phone number is insufficient evidence that a product order belongs to a particular exam attendee. An order from an outside prescription should not be forced into an in-house exam cohort. An existing optical customer newly attending an exam can be new to the clinical stream under your declared rule while remaining an established retail relationship.
If calculating the proportion of the defined exam cohort with a qualifying linked order, divide cohort members with that evidenced event by eligible members of the same cohort. Count people consistently and document exclusions. If the denominator is empty or linkage is unavailable, report unavailable. This identity defines a measure; it does not predict sales or justify a marketing budget.
Preserve independent retail acquisition
Keep a category for optical orders without an evidenced in-house examination link. It may contain outside-prescription customers, older relationships or missing records. Separate what is known from what is inferred. An absent link does not prove the customer is new, and a link does not prove the examination campaign caused the sale.
Give an optical campaign its own acceptance event when that is the purchased objective. Compare actual inquiries and orders with the agreed offer and full costs. Give examination acquisition its own relationship definition. This makes it possible to recognize useful retail work without overstating clinical growth.
Use the general measurement dictionary for upstream website events. The optometry dictionary adds the exam/order/fulfillment distinction and the outside-prescription route that general discovery metrics cannot resolve.
Frequently asked questions
Should optical sales be excluded from marketing reviews?
Keep them in a separate commercial series with the correct transaction definition. Excluding them entirely can hide useful work; combining them with new examination relationships obscures the acquisition question.
Can we use the same customer identifier everywhere?
Have the practice assess system meanings, permitted uses and matching reliability. A shared identifier does not remove privacy obligations or establish that the records describe the same business event.
What if our records cannot support a linked cohort?
Report separate examination and optical totals, state the gap and decide whether an approved prospective reporting change is worthwhile. Do not estimate a linked sales rate from unrelated totals.