Cost per lead benchmarks: how to think about them

A published healthcare cost per lead figure is context, not a target: it usually comes from another country, another definition of a lead and a mix of specialties. Question its source, sample and definition, then build your own benchmark as ranges by specialty, unit and intent, read alongside booked and honoured rates.

Every few months someone forwards me a chart with a healthcare cost per lead figure on it, usually with a note that says “are we paying too much?” Sometimes it comes from a board member, sometimes from an agency pitching for the account. The chart is almost always American, averaged across very different practices, and silent on what a “lead” means. The question behind it is fair. The chart is the wrong way to answer it.

I have argued before that cost per lead misleads when it is the only number. This piece is narrower and more practical: how to read a published benchmark critically, why it rarely transfers to an Indian specialty, and how to build an internal benchmark that actually helps you set budgets and judge an agency. It sits within the complete guide to Google Ads for doctors in India.

Why people want a healthcare cost per lead benchmark

The appeal is obvious. A single reference number lets a CEO judge a marketing team, lets a doctor in private practice judge an agency, and lets a CFO sanity-check a budget without learning how auctions work. In a function where so much feels unmeasurable, a benchmark looks like certainty.

The trouble is that cost per lead is a ratio of two things that vary enormously: what a click costs in your auction, and what share of clicks become whatever you call a lead. Both depend on the city, the specialty, the competition, the landing page, the desk and the definition. A benchmark collapses all of that into one figure from someone else’s accounts.

That does not make published benchmarks useless. It makes them context, not targets.

Five questions to ask of any published benchmark

Before a benchmark enters a board deck or an agency review, put it through these questions:

1. Where is the data from? Country, currency and period. An auction in one market tells you little about another.
2. Whose accounts? Benchmarks are usually built from the publisher’s own clients or tool users, which is not a random sample of advertisers.
3. What counts as a lead? Any tracked conversion, a form fill, a call of some length, a booked appointment? If it is not stated, assume the broadest definition.
4. Which category? A “health and medical” bucket can mix dentists, physiotherapists, fitness centres and surgeons, each with very different economics.
5. Average or median? A median resists outliers, an average does not. Neither tells you the spread, which is what you need to know whether your number is unusual.

A worked example. WordStream publishes an annual set of Google Ads benchmarks, and its 2026 edition is transparent about method: it describes the data as US-based search campaigns from April 2025 to March 2026, reports medians, states values in US dollars, covers Google Ads and Microsoft Ads, and includes categories such as dentists and physicians and surgeons. That is more disclosure than most. It still does not define a lead in a way you can map to your own funnel, and none of it is Indian. Useful for seeing how US healthcare auctions are moving; not a target for a clinic in Pune or a hospital in Indore.

Why benchmarks do not transfer to Indian specialties

Even an Indian benchmark, if you find one, needs care. The drivers of cost per lead in hospital ads vary more within India than most people expect.

• City and catchment. A metro auction with several hospital groups, aggregators and standalone clinics bidding on the same terms behaves nothing like a tier-2 city with two serious advertisers.
• Specialty and search intent. A routine consultation search, a second-opinion search and a planned procedure search attract different competitors, different click costs and different conversion rates, even within one department.
• Channel mix. Many Indian patients call or message on WhatsApp rather than fill a form. If your benchmark counts only forms, and your patients mostly call, the comparison is meaningless.
• Brand versus non-brand. Brand searches produce cheap leads from people who already chose you. Blending them into a specialty CPL flatters it.
• Language and geography targeting. Regional-language campaigns and wide radius targeting change both click cost and lead quality.
• Operational capacity. If the desk misses calls in the evening, your cost per lead rises for reasons no benchmark can see.

The honest summary: the only benchmark that transfers to your specialty in your city is one built from your own accounts.

When an external number is still useful

I do not throw published figures away. They are useful for direction rather than level. If a transparent benchmark series shows healthcare click costs rising year on year in its market, that is a reasonable prompt to check whether your own auctions show the same trend, and to budget for it.

They are also useful as a sense check on agency claims. If an agency promises a cost per lead far below anything you can find published anywhere, for a competitive specialty in a metro, ask how. The answer is usually brand traffic, a loose lead definition or leads that never book.

And they help frame questions for your own data. A benchmark that splits dentists from physicians is a reminder to split your own specialties, not a figure to hit. Used this way, an external chart earns its place in the appendix of a board pack. It should never sit on the front page as a target.

What a CPL in hospital ads hides

The companion piece on cost per honoured appointment covers the calculation that should sit next to CPL. The point here is narrower. Two campaigns with the same CPL can have very different value, because lead quality varies by keyword, by page and by hour of day.

A low CPL often means one of three things: broad keywords catching research and job searches, a form so easy that it collects casual enquiries, or brand traffic mixed into the number. A high CPL can be fine if the leads book and turn up. In accounts I have worked on, the specialty with the least flattering CPL has sometimes been the one whose leads most reliably became appointments, because its clicks came from precise, high-intent searches and its calls were answered by a trained coordinator who knew the doctors’ schedules. Cut that specialty’s budget on CPL alone and you cut the most dependable source of appointments in the account. That is why CPL should always be read with the lead-to-booked and booked-to-honoured rates beside it, and why your conversion tracking needs to reach those stages. The mechanics are in call tracking and offline conversion import for hospitals.

How to build your own healthcare cost per lead benchmark

An internal benchmark is a set of ranges, by segment, drawn from your own accounts and refreshed on a schedule. Here is the sequence I use:

1. Fix the lead definition. Write down exactly what counts: a submitted form, a call above an agreed duration, a WhatsApp conversation with a genuine enquiry. Apply it to every campaign and every unit.
2. Separate brand, non-brand and competitor traffic. Never report a blended specialty CPL. Your campaign structure by specialty and procedure should make this separation automatic.
3. Segment by specialty, unit and intent. Consultation, second opinion and procedure searches get separate rows even within one department.
4. Collect a full quarter before you call it a benchmark. Shorter periods catch seasonal swings, doctor leave and one-off events.
5. Record the funnel with it. For each segment: CPL, lead-to-booked rate, booked-to-honoured rate and cost per honoured appointment.
6. Express it as a range. A range with the median marked is more honest than a single figure, and it tells you when a week is genuinely unusual.
7. Annotate context. Note doctor changes, new competitors, landing page changes and tracking fixes, so a future reader knows why a number moved.
8. Refresh quarterly. Retire old ranges when the account, the market or the definition changes.

The table below is the structure I keep for each segment. The values come from your data; none are supplied here.

Setting a PPC budget for doctors from the appointment backwards

Once you know your funnel rates, you can set a CPL ceiling instead of borrowing one. Start with what the hospital can afford to spend to acquire one honoured appointment in that specialty, a figure finance should own, not marketing. Multiply it by your lead-to-honoured rate. The result is the highest CPL at which the campaign still pays for itself.

That arithmetic turns a vague question (“is our CPL good?”) into a specific one (“is this segment below its ceiling, and is the ceiling right?”). It also makes the budget conversation easier, because each specialty’s spend is tied to what its appointments are worth. The healthcare PPC budget and CPL calculator does this for you, and the wider method for building spend from service lines up is in building a hospital marketing budget from service lines.

Before you have history: estimates, not benchmarks

A new clinic or a new specialty has no internal data. That is where people reach for published numbers, and where I would reach for your own auction instead. Google’s Keyword Planner gives click cost and volume estimates for your actual keywords and locations, though Google’s own note on how Keyword Planner forecasts work is clear that they are projections, less reliable for new accounts, and not guarantees.

Treat the first weeks of spend as calibration. Run a tightly structured campaign on high-intent terms, with the keywords that book appointments rather than research terms, and with tracking to the booked stage from day one. What you learn in that period is worth more than any chart.

Presenting cost per lead to a board or a doctor

Boards and senior doctors do not need a lesson in auction dynamics. They need to know whether money is working. I present healthcare cost per lead in three lines per specialty: the current range against our own previous quarter, the lead-to-honoured rate beside it, and cost per honoured appointment against its ceiling.

If someone brings an external benchmark to the meeting, run it through the five questions in the room. It usually takes a minute to establish that it is from another country, another definition or another category. Then move the discussion to enquiry to appointment, the number that matters.

A short checklist before you accept any CPL number

• Is the lead definition written down and the same across segments?
• Is brand traffic excluded from specialty numbers?
• Is it a range from at least a quarter of your own data?
• Are booked and honoured rates reported beside it?
• Is there a ceiling tied to what an honoured appointment is worth?
• If it is external, do you know its country, sample, definition and category?

If any answer is no, the number is not ready to drive a decision. Fix the gap first, whether it is a definition, a missing funnel stage or a blended segment, and the benchmark conversation becomes short and useful.

Questions people ask