Sharing a specialty does not mean two clinics should share the same keywords, audiences or bidding strategy. Their strengths differ, and so do the customers who respond to them.
For a dermatology clinic in Gangnam, growly focused on finding the audience and bidding conditions that suited the practice. We aggregated response data, interpreted it and refined the campaign accordingly. The clinic and growly confirmed 24× ROAS for Month 5. This report explains the approach and the decisions behind iterative campaign improvement.
01 / A clinic-specific approach
We started with the clinic’s customers, not an industry template.
Dermatology campaigns often share similar keywords. Yet two people using the same search term may have different reasons for visiting, different timelines and different expectations. For international customers, planning a trip to Korea and looking for a clinic during a stay are distinct situations. A campaign copied solely on the basis of industry can miss those differences.
In this engagement, growly used customer response data to assess fit with the clinic and inform audience and bidding refinements. We did not treat the initial configuration as a finished answer. The data generated by each round of delivery became evidence for the next.
A suitable customer is more than someone likely to click. The relevant question is whether that interest fits the clinic’s services and can progress to a meaningful decision after an inquiry. That requires interpreting campaign response alongside the outcomes the clinic records.
02 / Reading the data
Read inquiry volume in its commercial context.
growly brought the clinic’s advertising data together to identify the audiences and delivery conditions that suited its business. The analysis focused on the difference between people who respond to an ad and people who choose the clinic’s services. A condition that produces many inquiries may have a different value when viewed against visits and revenue.
That assessment starts with clear definitions. An inquiry-button click, a conversation, a confirmed booking and an attended appointment are different events. A conversion in the advertising platform counts the action configured for measurement; it is not automatically a booking or a paying patient. Combining different stages changes how acquisition costs and advertising contribution appear.
Revenue also needs a consistent reporting period and basis. Advertising spend and the resulting visit or payment may occur at different times, and inquiries can lead to different revenue amounts. growly’s aim was to find conditions that attracted demand appropriate to this clinic. That understanding provides a basis for deciding where to expand and where to refine delivery.
03 / Audience refinement
Refine the conditions that fit this clinic.
growly reviewed the clinic’s campaign data and repeatedly refined its audience focus and delivery conditions. The approach connects responses across several dimensions with the clinic’s services, rather than treating one location or age group as a complete customer profile.
In the inquiry records for Month 5, entries recorded as US inquiries represented approximately 26%, the largest share for an individual country. Together with Taiwan and Singapore, the three countries accounted for about 43%. A large share of inquiries does not by itself identify the most efficient destination for advertising spend. Volume is a starting point for assessing opportunity; allocation also requires the cost of reaching that demand and the response after the inquiry.
More segmentation produces more comparisons but less data in each group. A small number of inquiries cannot establish that one group is consistently more valuable. Estimated income and interests cannot substitute for purchase intent. The table describes dimensions that may be reviewed across campaign types; it does not mean every option was used in this case.
Healthcare has additional limits on audience features and measurement data. Patient lists and sensitive medical-service purchase information cannot be treated as ordinary marketing data. Internal aggregate analysis and information transmitted to an advertising platform must remain distinct, with customer response assessed within the permitted scope.
| Dimension | Question for analysis | Availability distinction |
|---|---|---|
| Country, city and language | Where and in which language does demand appear? | Location and language are not nationality; distinguish presence from location interest |
| Age, gender and household income | Does the response difference persist? | Estimated categories include Unknown; income is available only in supported countries |
| Interests and purchase intent | Is this demand relevant to the service? | Campaign availability and sensitive-health restrictions apply |
| Parental and marital status | Does the category help explain customer choice? | Basic parental status: Display, Video and Demand Gen. Marital status belongs to detailed demographics |
| Custom segments | Do the selected signals describe relevant demand? | Not available in Search; additional limits apply to sensitive health ads and landing pages |
| Search terms, device and time | What are people seeking, and when do they respond? | Queries differ from configured keywords; available controls depend on the bidding strategy |
04 / The bidding balance
Assess click prices alongside customer acquisition value.
Through repeated refinement, growly sought a balance in bidding and delivery that suited this clinic. A lower click price is only part of the assessment. Low-cost clicks can still use budget poorly when they are unlikely to lead to bookings. Higher costs may warrant consideration when they offer access to demand that fits the clinic.
The important question is whether the current average performance can continue with the next budget increase. Additional spend does not necessarily buy the same quality of demand at the same price. The expected value of further delivery must be considered alongside the additional cost. A high average ROAS alone cannot establish how far to scale, just as a low CPC alone cannot establish that bids are well judged.
The controls depend on the bidding strategy. Manual CPC allows direct management of maximum click bids and supported adjustments. Smart Bidding sets bids at auction time around conversion goals; it does not simply add every manual location, time and demographic modifier. For this clinic, growly brought data together, identified suitable delivery conditions and continued to refine them in response to results.
05 / Reported results
Month 5 ROAS reached 24×. Recorded inquiries were 2.44 times the Month 1 level.
The clinic and growly confirmed an advertising revenue-to-spend ratio of 24× for Month 5. ROAS measures revenue relative to advertising spend. ROAS is not profit. It also does not isolate the additional revenue caused by advertising.
Summing the daily inquiry entries in the clinic’s dashboard by calendar month gives a Month 5 total approximately 2.44 times the Month 1 figure, an increase of about 144%. These are recorded clinic inquiries, reported separately from Google Ads-attributed conversions. We do not attribute the entire change to a single advertising adjustment.
The two measures answer different questions. ROAS describes the confirmed relationship between revenue and advertising spend. Recorded inquiries show the change in demand reaching the clinic. Our assessment considered both campaign response and the clinic’s records. Only finalized periods are included in this comparison.
Change in recorded clinic inquiries
Daily entries in the same dashboard are summed by calendar month. With Month 1 indexed to 100, Month 5 is approximately 244.
Month 1 = 100
- Month 1100
- Month 5244
Inquiry records, not deduplicated customers or Google Ads-attributed conversions. Index values are rounded.
06 / After a strong result
A strong result informs the next decision. It does not end the work.
This case did not produce a universal audience formula for dermatology. It built a body of evidence about one clinic and a way to use that evidence in campaign decisions. A different practice, even in the same neighborhood and specialty, may require a different answer.
As inquiries grow, the clinic’s response and visit processes also need attention. Do customers receive the information they need in time? Which questions recur in conversations? Has the customer mix or competitive environment changed? These questions help identify the next improvement and connect the opportunity created by advertising with the clinic’s results.
growly’s contribution is to carry those decisions through to delivery: read the response, identify the next adjustment and review the result. Time alone does not improve a campaign. Applying accumulated evidence to the next decision makes the work better suited to the business.
