Personal injury leads run $50 to $600 in most US markets. Shared leads sit at the bottom of that range, somewhere around $50 to $150. Meta lands between $150 and $300. Google Search runs $250 to $500, exclusive purchased leads $200 to $600, and live transfers go past $1,000 for the case types everyone wants.
That’s the answer, and on its own it’s close to useless.
What decides whether you make money isn’t the lead price. It’s cost per signed case, which usually lands somewhere between 7 and 20 times whatever you paid for the lead. Two firms can buy the same leads, from the same vendor, at the same price, in the same city, and finish the quarter three or four times apart on that number. The gap almost never comes from the ad account.
What you should expect to pay
| Channel | Typical cost per lead | What you’re actually buying |
|---|---|---|
| Shared purchased leads | $50–$150 | The same person, sold to three to five firms |
| Meta (Facebook, Instagram) | $150–$300 | Someone who wasn’t looking for a lawyer |
| Local Services Ads | $195–$400 | Billed whether or not you answer the phone |
| Exclusive purchased leads | $200–$600 | One firm’s access, on the vendor’s word |
| Google Search | $250–$500 | Someone searching for you right now |
| Live transfers | $500–$1,500+ | A pre-screened caller already on the line |
Add 20 to 40 percent if you’re in Los Angeles, New York, Miami, Houston, or anywhere comparable. Case type moves it again. Reported Meta costs come in around $150 to $250 for auto, $200 to $300 for slip and fall, $250 to $400 for med mal, and $300 to $500 for wrongful death.
Use those ranges the way you’d use a Kelley Blue Book number. They tell you when someone’s quote is nowhere near reality. If a vendor offers you $80 exclusive MVA leads in Miami, one of those three words isn’t doing the job you think it is.
Why the number your agency quoted is probably wrong
Nearly every “2026 benchmark” floating around legal marketing is older or thinner than it looks, and the one you’ve most likely been shown is both.
You’ve seen the table. Google Search $442, Facebook $286, LSA $378, SEO $183. It’s everywhere. It comes from a First Page Sage study, and if you read the methodology note instead of the table, you’ll find the campaigns behind those numbers ran from January 2022 through December 2024. The study says so plainly. The agency blogs recycling it don’t.
The disagreement between sources is worse. Google Search cost per lead for personal injury has been published at $131.63, $284, $325, $442, and $500 to $1,500, all of it dated 2026. That’s a tenfold spread on one metric in one practice area.
The reason matters, because it’s the same reason your own monthly report might be lying to you. Nobody defines what a lead is. One source counts every form fill including the bots. Another counts any call that ran longer than thirty seconds. A third counts only the inquiries intake bothered to write up. Same campaign, three completely different numbers, and none of them mention which rule they used.
The most honest source I’ve found is WordStream’s 2026 report, mostly because it publishes its method: 13,474 US campaigns, April 2025 through March 2026, reported as medians rather than means. It puts attorneys and legal services at $9.87 a click and $131.63 a lead, the most expensive of the 23 industries they track, converting at 5.55 percent. Treat that as a floor. It blends divorce, bankruptcy, and people googling a firm by name in with injury work, and PI sits well above the blend.
Cost per signed case is the number that matters
Cost per signed case is your total marketing spend divided by the matters you actually retained in the same period. Everything else is a proxy.
Here’s the same $250 lead run through five different sign rates:
| Sign rate | Cost per signed case |
|---|---|
| 5% | $5,000 |
| 10% | $2,500 |
| 15% | $1,667 |
| 25% | $1,000 |
| 40% | $625 |
Nothing about the campaign changed between those rows. Same ads, same price, same market. The entire spread comes from what happens after the lead lands, which means the firm signing at 5 percent and the firm signing at 25 percent are running an identical media buy and operating two different businesses. Usually only one of them thinks the problem is marketing.
For the full formula, including the attribution lag that trips up most monthly reports, see how to calculate cost per signed case.
What this looked like in our own account
We run MVA campaigns for a general-practice firm in Montana — criminal, family law, and personal injury under one roof. I’m going to give you the whole account, including the part that makes it look bad, because the bad part is the useful part.
Over the life of the campaign we spent $5,158 and got 9 website leads. Blended out, that’s $573 a lead, which is a rough number by any standard. It’s also not the whole story. A payment card on the account failed and the campaign went dark for a stretch, and we ate two learning-phase resets that sent Meta back to square one on optimization. Once the leading campaign settled, cost per lead came down to $242. Recent leads have been running around $269.
That spread, $573 blended against $242 once it stabilized, is the most instructive thing in the file. It’s the difference between what a firm sees if it judges a campaign at week three and what the same campaign costs once it’s left alone long enough to learn.
Of those 9 leads, 4 or 5 were qualified enough to bill. Two signed. Those two produced $17,000 in attorney fees.
| Measure | Figure |
|---|---|
| Total media spend | $5,158 |
| Website leads | 9 |
| Blended cost per lead | $573 |
| Stabilized cost per lead | $242 |
| Qualified leads | 4–5 |
| Signed cases | 2 |
| Cost per signed case | $2,579 |
| Attorney fees generated | $17,000 |
| Return on media | 3.3x |
Now the caveats, because they’re real. Nine leads and two cases is one account, not a dataset. Swap a single outcome and the sign rate moves ten points, so read the percentages as direction and the dollars as fact. And the $2,579 is calculated across total spend, dead period included. If I ran it on the stabilized weeks alone the number would look a good deal better, and it would also be exactly the kind of hand-picked denominator this article is telling you to distrust.
What I’ll stand behind: media consumed about 30 percent of the fees it generated, and the campaign returned 3.3 times what went into it. Those two numbers only exist because the firm owns its ad account and somebody bothered to trace the signed cases back to the leads that produced them. Most firms can’t build this table for their own campaigns, and that’s the part worth sitting with.
That $2,579 gets a rough second check from a different cut of the same firm’s numbers. In July, the firm spent $12,537 across every channel it runs, Google Ads, LSA, Meta, and SEO, and signed 20 new cases across all three practice areas. Personal injury’s slice of that spend was $3,000 on Meta plus roughly $493 of the LSA budget, since LSA is a Criminal-and-PI blend that runs about 90 percent Criminal. Media only, that’s $3,493 against 2 signed PI cases, or $1,746 a case. A different month, a wider set of channels, and it lands in the same range as the cumulative Meta-only figure above.
One honest complication: as of the numbers I’m working from, 6 of those 8 July PI calls are still sitting at outcome unknown. If two more sign, cost per case drops by roughly a third. Small accounts move like that, and it’s the reason nobody should be handed a single month’s number and told it’s settled.
None of which is an argument for chasing a lower cost per lead. Pulling CPL down 20 percent is genuinely hard and takes real media skill. Getting a sign rate from 10 to 15 percent is usually a staffing and response-time problem, and it beats the media work outright.
Sign rate moves on things nobody puts in a benchmark. How fast anyone picks up, including at 9pm on a Saturday. Whether intake makes six follow-up attempts or gives up after two. Whether a lead gets a real sequence or a single voicemail and a shrug. Whether the person taking the call qualifies for liability and injury or just writes down a name and a number. Whether you’ll even take the case types your ads are bringing in.
All of that costs more than the media does.
What realistic looks like
Most firms get sold something considerably rosier than this, so here it is straight.
Month one isn’t the number. Meta and Google both need conversion volume before optimization stabilizes, which means your early cost per lead has almost no relationship to your steady-state cost per lead. Give a campaign 90 days before you judge it, and judge it on signed cases.
A rising cost per lead is sometimes the good outcome. Move a campaign from $40 unqualified form fills to $180 qualified phone conversations and the dashboard says you got 4.5 times worse. The business usually got better. If your cost per case dropped while your cost per lead climbed, that campaign improved, and anyone who can’t walk you through why shouldn’t be running your account.
Nobody can promise you case volume. An agency guaranteeing signed cases is either buying leads and reselling them to you at a markup, or building toward an argument about attribution in month four. Sometimes both.
The first six months cost more per case than the steady state. Pixel history, conversion volume, and audience data all compound. A firm that owns its ad account gets a cheaper case in month twelve than it did in month two on identical spend. A firm buying leads pays the same rate forever, because the asset that improves belongs to the vendor.
And twelve months of your own numbers will beat every benchmark published this year. Your market, your case mix, your intake desk. Nothing else describes your firm.
Five questions to ask before you sign anything
Run these on whoever you’re working with now. Ten minutes, and you’ll know where you stand.
1. What’s my cost per signed case, by channel, for the last 90 days?
Not cost per lead. Not cost per conversion. Signed cases. If what comes back is a cost per lead, nobody is measuring the thing that determines whether you’re profitable.
2. How do you define a lead?
Any consistent answer is fine. What you’re listening for is whether an answer exists at all, and whether it’s the same one they gave last quarter. If the definition moves, every number you’ve ever been handed is floating.
3. Does my signed case data get sent back to Google and Meta?
Ad platforms optimize toward whatever you feed them. Feed them form fills and they’ll spend the next year finding you cheaper form fills. Feed them signed cases, through offline conversion uploads or the Conversions API, and they start hunting for people who retain. Most law firm accounts have never had this wired up, which is why so many campaigns produce impressive lead volume and unimpressive case volume.
4. Who owns the ad account, the pixel, and the audiences?
If the agency’s Business Manager owns the ad account and you’re a user on it, you own nothing. The pixel history, the conversion data, the lookalike audiences you paid to build, all of it stays behind when the relationship ends. Paying the invoice every month doesn’t establish ownership, and a surprising number of firms find that out the hard way.
5. If I leave in six months, what do I walk away with?
Ask for the list in writing. Ad account, Page, pixel or dataset, custom audiences, creative files, campaign history, call recordings, CRM export. Get it settled while everyone still likes each other.
What about leads from ChatGPT and AI search?
Nobody can price this honestly yet, and I’d be careful with anyone who says otherwise.
Injury victims are asking AI assistants what to do after a crash and who to call in their city. That behavior is real and it’s growing. The measurement, though, is broken by design.
Start with the fact that most citations never produce a click. SparkToro research from January 2026 put Perplexity’s citation-to-visit rate at 12 to 18 percent, meaning four times out of five your firm gets named in front of somebody who needs a lawyer and nothing shows up anywhere. Google’s AI Mode strips the referrer deliberately, so that traffic is invisible to standard analytics. And the biggest category doesn’t register at all: somebody reads your firm’s name in an AI answer, googles you three days later, and lands in your reports as branded search.
There’s one published figure out there, $246 a lead, and it rests on a single year of 2024 data by the study’s own admission. That’s a first estimate from the channel’s earliest months, not a benchmark.
You’ll also run into the claim that AI traffic converts at 4.4 times organic. Think about who’s in that number. Someone who reads an AI summary, looks at the cited sources, and then clicks through has already done most of their own filtering. The channel isn’t converting better, it’s qualifying earlier, and the 80-odd percent of citations that never produced a click aren’t in the denominator.
So don’t chase a cost per lead here. Put a real open-text question at intake asking how the person found you, watch your branded search volume as a lagging signal, and treat AI search the way firms treated billboards for fifty years. It’s a demand-shaping channel being measured with demand-capture tools. Investing without a CPL is defensible. Claiming you have one isn’t.
Frequently asked questions
How much does a personal injury lead cost? Between $50 and $600 in most US markets. Shared purchased leads run $50 to $150, Meta $150 to $300, Local Services Ads $195 to $400, exclusive purchased leads $200 to $600, and Google Search $250 to $500. Live transfers exceed $1,000. Add 20 to 40 percent in major metros.
What’s a good cost per lead for a personal injury firm? There’s no universal answer, because it depends entirely on your sign rate and your average fee. A $500 lead that signs at 20 percent produces cheaper cases than a $150 lead signing at 4 percent. Judge cost per signed case instead.
What is cost per signed case? Total marketing spend divided by the matters you actually retained in the same period. In personal injury it typically runs 7 to 20 times the cost per lead, and the spread comes down to intake performance far more than media performance.
Are Facebook leads cheaper than Google leads for PI firms? Per lead, yes, usually by a wide margin. Per signed case, often not. Meta reaches people who weren’t searching for a lawyer, so more of those leads won’t qualify. The comparison only means something once you filter for qualified inquiries.
Why do published PI lead cost benchmarks disagree so much? Because nobody defines what a lead is. One study counts raw form fills, another counts calls over thirty seconds, another counts only qualified inquiries. Vintage is the second problem: several widely quoted 2026 figures are built on campaign data from 2022 through 2024.
Is buying leads cheaper than running your own campaigns? Cheaper to start, more expensive to sustain. Purchased leads carry no setup cost and no learning period. Running your own costs more in months one through three and less after that, and it leaves you owning the ad account, the pixel history, and the conversion data. Your time horizon decides which one wins.
What does a lead from ChatGPT or AI search cost? Nobody can answer that honestly yet. Most AI citations never produce a click, Google AI Mode strips the referrer by design, and most AI-influenced visits arrive as direct or branded search. The one published figure rests on a single year of 2024 data. Track it through intake self-reporting and branded search trends instead.
The short version
Lead price is negotiable and mostly beside the point. Cost per signed case isn’t.
Most firms convinced they have a lead cost problem are actually looking at a sign rate problem, an attribution problem, or an ownership problem, and all three are cheaper to fix than a media buy. If questions two through five gave you trouble, start there.
That diagnosis is the work we do. Here’s how we approach personal injury case acquisition.
Sources
- WordStream by LocaliQ, 2026 Google Ads Benchmarks. 13,474 US campaigns, April 2025 to March 2026, reported as medians.
- First Page Sage, Average Personal Injury Cost Per Lead. 49 firms, campaigns from January 2022 to December 2024.
- Pareto Legal, Meta Ads for Personal Injury Lawyers. Agency book of business, 2026.
- Rankings.io, How Much Do Personal Injury Leads Cost. Purchased lead pricing by case type.
- Foundry CRO, Tracking AI Search Referrals. Carries the SparkToro January 2026 Perplexity finding.
Last reviewed August 13, 2026. Refreshed quarterly.