An advertising benchmark is useful only when its definitions and context match your practice. Cost per click, cost per inquiry and cost per attended consultation answer different questions. Before comparing your campaign with an industry average, build a target from the economics and capacity of the service you are promoting.
Define the outcome worth purchasing
Choose whether the working target is a valid inquiry, a booked consultation or an attended visit. The further downstream the event, the more it reflects the whole operating process. Keep intermediate measures so you can diagnose problems, but do not label every button click as a new patient.
Work backward with explicit assumptions
Suppose a hypothetical practice sets a media-only ceiling of $200 per attended consultation. If 25% of valid inquiries attend, the corresponding inquiry ceiling is $50. If 5% of ad clicks become valid inquiries, the implied click ceiling is $2.50. These are planning assumptions, not market benchmarks or client results. Change any assumption and the allowable costs change.
Add the costs outside the auction
Campaign management, creative production, landing-page work and staff time may affect whether the program is worthwhile. Keep media-only and fully loaded views separate and clearly named. Do not compare one provider's all-in cost with another provider's advertising spend alone.
Investigate variation before reacting
Service mix, geography, competition and inquiry handling can all affect results. Review the search terms and destination pages, then follow the inquiry cohort through the office. A higher cost per inquiry may still produce a better attended-visit cost if the inquiries are more appropriate. Conversely, cheap contacts that cannot be reached may offer little value.
Set a review rule
Define a spending ceiling, a minimum observation period suited to the decision cycle and a process for investigating poor results. Do not keep spending solely because the campaign has not reached an arbitrary benchmark. When data is sparse, report uncertainty and use a controlled next test. Your target should be a management tool that can be revised when real evidence replaces assumptions.
