What paid ads cost a high-ticket practice, and when the math clears

Paid ads for a high-ticket service either clear or burn. The difference is arithmetic, and it takes about ten minutes to run.

Ads work in one narrow sense. They deliver clicks at a market price.

Whether you can pay that price comes down to two numbers. The first is what a client is worth to you. The second is how many clicks it takes to get one.

So start with what the market charges.

What a click costs a professional practice right now

The most useful cost benchmarks come from WordStream and LocaliQ. They publish medians each year, drawn from their own client accounts. The 2026 report covers 13,474 US search campaigns run between April 2025 and March 2026.

In the Business Services category, the median click costs $5.87. The median conversion rate is about 5%. The median lead costs $94.

Attorneys and legal services pay more. Their median click is $9.87 and their median lead is $132. That is the highest lead cost among the 23 categories tracked.

Two things about those numbers matter more than the numbers.

A lead in that data is a form fill or a phone call. It is not a client. For a practice selling a $20,000 engagement, the gap between the two is the whole question.

A median also hides its own top end. Half of legal keywords cost more than $9.87. The terms carrying the most buying intent sit well above the middle. Your own terms have their own price, and the keyword planner will show it before you spend anything.

One more thing is worth knowing about the benchmark corpus. Almost everyone publishing cost data on paid search sells campaign management. That does not make the numbers wrong. It does mean the framing around them leans optimistic. That is a reason to run your own math.

The number that decides it is the cost of a signed client

Take a consultant whose average engagement is worth $20,000.

At the category median, a lead costs $94. Suppose one in four leads becomes a consultation. Suppose one in three consultations becomes a client. That makes 12 leads per client, so the client costs about $1,100 in ad spend. On a $20,000 engagement, that spend is roughly 6% of revenue. It clears with room to spare.

Now change two inputs. The practice bids on broad category terms in a competitive metro, and clicks cost $25. The ads point at the homepage, which converts at 2%. That page was built to introduce a firm, and capturing inquiries was never its job.

A lead now costs $1,250. At the same close rates, a client costs $15,000. On a $20,000 engagement, that is the bonfire.

Two inputs moved, and the campaign flipped from profitable to ruinous. The market did not change. The service did not change. The price of the click and the conversion rate of the page did.

Work backward from the client to your maximum bid

The cost-per-client calculation is more useful run in reverse. That way it produces a number you can check against the auction today.

The reverse calculation takes three inputs, and you either know each one or can estimate it closely:

• What a new client is worth in the first engagement, before any repeat work.
• What share of that you will spend to acquire one. Ten percent is a common ceiling in professional services.
• How many inquiries it took to sign each client over your last twelve months.

Start with a $20,000 engagement and a 10% ceiling. You can pay $2,000 to acquire a client. If 12 inquiries produce one client, your maximum cost per inquiry is $167. If the page behind your ad converts at 5%, your maximum bid is $8.35 a click.

That maximum bid is the figure to compare against what your terms cost. When the auction sits below it, paid search is a purchase. When the auction sits above it, the campaign runs at a loss. Good management narrows that gap without closing it.

Many owners have never seen this number. It is why the conversation with an agency tends to start at budget.

At the median, the big paid channels return less than they cost

Dreamdata builds attribution software for business-to-business companies. Its yearly benchmark report draws on 66 million sessions and 3.5 million customer journeys. The 2026 edition measures closed-won deals across a twelve-month window.

LinkedIn's median return on ad spend came in at 121%. Google Search came in at 67%. Meta came in at 51%.

Read the two lower figures again. At the median, a dollar spent on paid search returned 67 cents of attributed revenue inside the year.

Three caveats belong with that finding. Dreamdata's customers skew toward software companies with long sales cycles. The report is a LinkedIn benchmark study, published by a vendor that sells measurement. And a twelve-month window on first deals leaves out the repeat work and referrals a signed client brings later. For a professional practice, that later value can be most of the value.

In that dataset, more than half of the accounts running paid search returned less than they spent. The caveats change how far the finding travels, and they leave that fact standing. Run the same test on every channel you fund, and the ranking sorts itself.

Three conditions have to hold at once

The first condition is value per client, and most expert practices satisfy it. A $20,000 engagement absorbs a $1,100 acquisition cost without strain. A $2,000 project does not.

The second condition is targeting narrow enough that most clicks come from buyers. The phrase someone types decides this. A person searching "succession planning attorney Boston" is shopping. A person searching "succession planning" is reading. You pay the same for both.

The third condition is the page behind the click. It does the work of a first meeting, and a homepage rarely does that work. Drop a page from 5% to 2% and every other number in the campaign more than doubles.

The common failure is two out of three. Valuable clients and decent targeting sit behind a page that leaks the traffic. On the monthly report it reads as a cost-per-lead problem. The fix sits on the landing page.

For an expert practice, the cheapest paid buy is usually your own name

The standard playbook skips the cheapest paid buy an expert practice has.

When someone hears your name, they search it. Anyone can bid on that name as a keyword, including the firm across town. Google does not restrict trademarks in keyword targeting. What a competitor may not do is put your name inside their ad text.

So the results page for your own name can carry a rival's ad above your website. You would not know unless you looked.

Defending that page is inexpensive. Branded terms draw few competing bidders. Your relevance score on your own name is high. The clicks also convert better than anything else you could buy, because the person already knows who you are.

The second cheapest buy is the people who visited your site and left. They arrived for a reason. Showing them a modest number of ads over the following weeks costs a fraction of prospecting to strangers.

Both are small campaigns. Neither fills a pipeline on its own. An agency proposing a five-figure monthly budget will find them unexciting. For a referral-fed practice, they are often the only paid spend that clears in the first year.

Where the spend burns

Broad category terms are the first place the money goes. They bring volume, and volume reads as progress. Most of that volume is people reading rather than hiring.

Automated matching is the second place. Google's Search campaigns now expand past the keywords you chose, and the newer formats expand further. The system finds you cheap conversions, because cheap conversions are what you asked for.

Negative keyword lists and a weekly look at the search terms report are the controls. Both used to be tedious enough that small advertisers skipped them. That excuse expired. Export the report, hand it to ChatGPT or Claude, and ask which queries spent money without producing an inquiry. The review takes minutes.

The conversion event is the third place. Tell the platform that a form fill is the goal, and it finds people who fill in forms. Import the signed client back from your own records, and it optimizes toward people who sign. On a small account, that one change does more for lead quality than any bidding tactic.

The fourth place is a market too small to hold an auction. Some expert niches draw a few dozen searches a month nationally. No bid strategy fixes an empty auction. The keyword planner tells you this for free in about five minutes.

Size the test to your sales cycle

Business buying is slow. The same Dreamdata report puts the average journey at 272 days. About 220 of those days pass before the buyer contacts a seller.

A 30-day test on a nine-month cycle measures almost nothing. It shows clicks, form fills, and a cost per lead. None of those say whether a client arrived.

So a fair test runs long enough to sign someone. The other option is to scope it to what it can measure. That means the cost of an inquiry, and whether the inquiries look like the people you want.

Set the maximum bid you calculated as the ceiling. Set the total you are willing to lose before you start. Then hold both.

Whether to fund ads at all, against the slower work that compounds, is its own question.

The arithmetic is the answer

Paid advertising does not create demand for expertise. It buys access to demand that already exists. The price is set by everyone else who wants the same buyer. Some of them can pay more than you can.

So the answer to "do ads work for a practice like mine" is rarely yes or no. It is a number: the most you can pay for a click and still make money.

Every practice has that number. Most owners have never worked it out. The ones who have can tell in ten minutes whether the auction is worth entering.