If your cost per signed case is uncomfortable, the click price is almost never the reason. A personal injury firm paying $95 a click and a firm paying $140 a click can end up at the same cost per case, and the cheaper one is often the one doing worse. What separates them is everything that happens between the click and the retainer.
So work the chain backwards, because that is where the leaks are.
The arithmetic you should be running every month
Take a search campaign at $80 a click. Say 5% of clicks produce a call or form fill. That is one contact for every 20 clicks, so $1,600 a contact. Now apply the rest of the chain:
- 70% of calls get answered live — $2,286 per answered contact
- 40% of answered contacts are in your practice area, in your jurisdiction and inside the statute — $5,714 per genuine prospect
- 35% of genuine prospects sign — $16,300 per signed case
None of those percentages are invented on your behalf. Pull yours from the phone system and the case management software and do the multiplication. Do it before you touch a bid.
Now change one number. Answer 90% of calls instead of 70% and the same spend produces a case at about $12,700. You did not buy a cheaper click. You bought better intake cover, and it cut cost per case by more than a fifth.
That is the whole argument. The click is the smallest lever in the model and the only one most firms pull.
Intake is the largest single lever, and it is not a marketing line item
Paid search sells urgency. Somebody searching "car accident lawyer near me" at 9pm on a Sunday is not comparison shopping over a fortnight — they are calling down the ad results until a human picks up. If you are the fourth call and the first three answered, you paid $80 to fund somebody else's file.
What to check, in order:
- Missed and abandoned call rate by hour and day. CallRail and CallTrackingMetrics both break this out. Most firms are surprised by evenings, Fridays after 4pm, and the whole weekend.
- Speed to first contact on form fills. Under five minutes and you are talking to somebody who still remembers filling the form. An hour later you are leaving a voicemail for a person who has already signed elsewhere.
- Who answers. A receptionist reading a script and a trained intake specialist produce different sign rates on identical traffic.
The fix is often cheap. An after-hours legal answering service — Alert Communications, Smith.ai, Ruby and similar — runs roughly $1.50 to $3.50 a minute, or a few hundred a month on a small plan. Against a $16,000 cost per case, one recovered signing pays for a year of it. If you would rather keep it in house, a routing and follow-up layer in something like GoHighLevel that fires an SMS the moment a form lands and rings three mobiles in sequence costs a fraction of what the missed calls do.
Before you increase ad spend by a dollar, listen to twenty recorded calls. It is the least enjoyable two hours in this job and the most profitable.
You are paying for cases you would never take
Broad match plus automated bidding will happily buy volume that has nothing to do with your book. The usual offenders:
- People looking for free advice, legal aid or a pro bono referral
- Other lawyers, paralegals and students researching
- Wrong jurisdiction — an out-of-state searcher on a mobile network routed through the wrong city
- Right practice area, no case: no injury, no treatment, no coverage, or past the statute
- Existing clients calling the ad number because it is the one they saw
Sort your calls into buckets and tag them in whatever you use for intake — Lawmatics, Clio Grow, Filevine or Litify — then push signed cases back into Google Ads as an offline conversion import. That single piece of plumbing changes what the bidding algorithm is chasing. Optimising to form fills tells Google to find you more people who fill in forms, and it is very good at that. Feed it retained cases with values attached and it starts hunting the queries that produce them. Building that pipeline is the practical half of conversion tracking and reporting, and it matters more in legal than in almost any other trade because the ratio of contacts to cases is so brutal.
Negatives and geography: the cheapest hour in the account
Pull the search terms report for the last 90 days, sort by cost, and read every term that spent more than one click's worth without producing a contact. Expect to find "free", "how to", "salary", "jobs", "public defender", "reddit", "template", "attorney near me" in a city you do not serve, and a long tail of competitor names you may or may not want to keep.
Geography deserves as much attention. Two settings people get wrong:
- Presence versus interest. Google's default has historically included people merely showing interest in your area. If you only take cases in one state, set location options to presence only, or you will pay for browsers three time zones away.
- Radius targeting across a metro. A 25-mile radius from a downtown office in Houston or Chicago covers very different case values at each edge. Split the radius into tiers and bid them separately rather than paying one price across the whole map.
The same logic applies to practice areas. Personal injury and family law should almost never share a campaign — the click prices differ by a factor of five and the intake scripts have nothing in common.
Why adding budget usually makes cost per case worse
This is the part firms find counterintuitive. The numbers look acceptable, so you double the budget and expect roughly double the cases. Three things happen instead.
You buy further down the quality curve. Your existing spend was already taking the best-matched impressions available. Extra budget buys looser matches, broader queries and worse positions on the good ones. Marginal cost per case is always higher than average cost per case, sometimes by a lot.
You bid against yourself in a thin market. Legal auctions in a single metro have a finite number of high-intent searches a day. Raising bids where three firms are already at the ceiling raises everybody's cost and moves very little share.
Your intake breaks first. A team that comfortably handles 40 enquiries a month does not handle 80 to the same standard. Answer rate falls, follow-up slips, and the conversion rate drops across the whole volume — including the leads you were already winning. The cost increase applies retroactively to traffic that used to work.
If you are going to scale, scale intake capacity first and the budget second. Run the extra spend as a separate campaign so you can see what the marginal money actually bought.
What this should cost
Bands, not quotes. Everything here moves with market, practice area and how many firms are bidding in your county.
| Item | Typical range | What moves it |
|---|---|---|
| Click cost, personal injury | $60–$400+ | Metro size, case type, competitor count. Truck accident and mass tort sit at the top. |
| Click cost, family law | $12–$60 | Contested divorce and custody bid higher than uncontested filings. |
| Click cost, estate planning or immigration | $8–$45 | Wider intent spread, more informational traffic to filter out. |
| Agency management fee | 10–20% of spend, or a flat monthly retainer | Account complexity, number of practice areas and markets, whether landing pages and tracking are included. |
| Sensible starting ad budget, one practice area, one metro | $3,000–$10,000 a month | Below the bottom of that band in personal injury you will not gather enough data to make decisions with. |
Our own PPC management starts from $200 a month as a minimum, and what a given firm pays depends on how many campaigns, markets and landing pages are involved. The pricing page sets out how that is built up, and the ROI calculator lets you sanity-check a budget against your own average case value before you speak to anybody.
When paid search is the wrong buy
Worth saying out loud, because plenty of firms are sold PPC when something cheaper would serve them better.
- Your average case value is under about $2,500 and your market is competitive. The arithmetic at the top of this article stops working. Put the money into local SEO and the Google Business Profile instead.
- Nobody can answer the phone outside office hours and you will not pay for cover. You will be funding your competitors. Fix intake first.
- Your site takes six seconds to load on a phone. Paid traffic is mostly mobile and mostly impatient. A page rebuild is a better first cheque than a campaign — that is conversion rate optimisation work, not media work.
- You have never tracked a call. Spend the $50 a month on call tracking, run it for a quarter, and make the budget decision with data.
If the model does work for you, the order of operations is: fix intake, tag outcomes, feed signed cases back to the platform, tighten geography and negatives, then look at the bids. Most firms work that list in reverse and wonder why the cost per case will not move.
If you want a second pair of eyes on the account and the intake chain together, that is the work we do for law firms. Book a call and bring the last 90 days of search terms with you.
