A household making more than 100 round trips a year picks the centre it can sustain, not the one with the better name. A centre 8 km away beats a better-known one 30 km away. An enquiry from 22 km outside the catchment is not a lead, and a city-level report counts it as one.
Dialysis is a capacity business. No media spend adds a chair to a centre running at capacity, so every rupee pointed at it buys enquiries you have to decline. The reverse costs more: an empty chair accrues rent, depreciation and a rostered technician every day, whether anyone sits in it or not.
Much of the placement is done by a referring clinician or a discharge desk. That audience is not moved by creative. It is moved by whether a slot exists this week, whether the shift timing suits the patient’s household, and whether someone answers the phone. Reliability is the pitch.
Open a site 6 km from an existing one and overlapping catchments can divide a single demand pool across two rent bills. Network enquiry volume, cost per enquiry and patient count can all look unchanged for months while both centres run at partial utilisation carrying full fixed cost.
Every part of this programme is organised around a physical location and the chairs inside it. The unit of planning is the centre and the shift, and the budget is judged against utilisation rather than against enquiry volume.
We define each catchment in travel time at the hour the shift actually runs, then plot the residential pin codes of current patients over it. Networks already hold that data and rarely map it. The gaps are the addressable demand, and they usually sit in a different direction than the media plan assumed.
Radius-capped local search and maps placement around under-filled centres, throttled by shift and reviewed weekly. What we do not run is metro-wide awareness and broad social reach, because most of that delivery lands on households that could never sustain the journey, and the fraction that converts near a full centre produces an enquiry you have to turn away.
One page per centre carrying what decides the choice: exact location and landmarks, shift timings, current availability, insurance and payment arrangements, and how to reach a named human. A single Dialysis Services page cannot answer a query anchored to a locality, and a network with 12 centres has 12 markets.
Assistants now field the nearest-centre question directly. Generative engine optimisation for a dialysis network is mostly an accuracy problem: entity, address, hours and availability data that stay consistent across your site, your listings and the sources these systems read, so the answer names the right building.
A live referrer list per centre showing who sent patients this quarter and who stopped, open slots pushed out before anyone asks for them, and a named person to reach. It is unglamorous, it is rarely staffed, and it moves more placement than any campaign in this specialty.
Patients stop attending because the household relocated, because a clinician moved them to a different centre or treatment, or because the patient died. Only one of those is a marketing signal, so exits are separated before anything is reported as churn, automated messaging has to be capable of stopping immediately, and the retention metric becomes chair-months served rather than a churn percentage.
The consideration window here is short and the relationship is long, which is the reverse of most healthcare marketing. Placement is usually settled within days of a nephrologist’s instruction or a discharge desk handing over a slip, and it is often settled by the clinician rather than by the household. After that the same family travels to the same building several times a week for as long as they live in that neighbourhood. Nothing about that pattern is served by a metro-wide campaign optimised for cheap enquiries. We plan per centre, per shift, against utilisation.
Travel-time catchments per centre per shift, overlaid with where your current patients live, to locate real addressable demand.
A page and a verified profile per centre, built around location, timings, availability and a named contact.
Consistent entity, address and availability data so AI answers to nearest-centre questions name the correct building.
Radius-capped local campaigns allocated against empty chairs, throttled by shift and reallocated weekly.
A named referrer layer per centre with slot availability pushed out, activity tracked and a review cadence that matches the media.
Cost per patient acquired against chair utilisation, chair-months served, and exit reasons separated before churn is reported.
Ad platforms treat health status as a sensitive category, so interest targeting, custom audiences and lookalikes built on a medical condition are not available. Targeting has to be geographic and intent-led instead, which is one reason catchment accuracy carries the weight that audience strategy carries in other categories.
In a category decided by proximity, the map surface is the channel. Each centre needs its own verified profile with hours that match the shifts actually staffed. A missing listing, a merged one, or hours that disagree with reality removes a centre from the only placement that reliably converts, and none of that sits inside a media dashboard.
Comparative clinical claims are not permissible in marketing, and patient testimonials drawn from a long-term chronic population carry consent and privacy exposure that is not worth the conversion lift. Credibility here is built from published operational detail: timings, availability, staffing, cost structure and who to call.
The cheapest enquiries come from the widest catchment, so optimising for cost per lead actively pushes spend towards households that can never attend and towards centres that cannot accept anyone. The number improves while the business does not.
Our healthcare work spans hospitals, clinics, home care and clinician brands, including Apollo Home Healthcare. What we bring to a dialysis network is a method built from its economics: a fixed number of chairs, a market measured in travel time, and a budget that has to answer to utilisation rather than to reach.
The binding constraint is the chair, so the plan starts from utilisation by centre by shift and works backwards to media. Spend is throttled where centres run full and pushed where chairs sit empty, on a weekly cadence rather than a monthly one.
Every centre gets its own search estate, its own listing hygiene and its own demand map. That is more work than a single service page and it is the only structure that matches how families and assistants actually ask the question.
Cost per patient acquired sits directly beside chair utilisation, split by centre and shift, with exits separated by reason before anything is read as churn. The first month those columns disagree, you find out which one has been running your marketing.
upGrowth runs acquisition for Apollo Home Healthcare, so a service delivered repeatedly to the same household over a long period, judged on continuity rather than on a single conversion, is territory we work in directly rather than theoretically.
The metric is cost per patient acquired against chair utilisation, not cost per lead. A dialysis network can post falling enquiry costs across a quarter in which one centre turned people away on the morning shift and another ran short every day, because a city-level average has no meaning in a business made of drive-time radii.
What follows from that is commercial, not analytical. If utilisation is the target, the correct recommendation for a centre at capacity is to cut its spend to zero and move the money 9 km across town. No agency paid a percentage of media will ever make that call, and no scope written around monthly enquiry volume will let them. So the engagement has to be structured for it: a fee that does not rise with spend, reallocation authority at centre and shift level, a utilisation target agreed with operations, and overlap modelling delivered before a new lease is signed rather than after the second centre underperforms.
The mechanism behind all of this is set out in Pain in the Kidney, our field note on why dialysis acquisition is a catchment problem rather than a messaging one.