A marketing director at a 19-attorney PI firm showed us her tracking setup last quarter: a Google Sheet with 9 vendor tabs, a column labeled “ROI (estimated),” and a sticky note on her monitor that read “confirm settlement numbers with intake.” She wasn't behind the curve. For most PI firms, that spreadsheet is the state of the art.
Her explanation was one we hear constantly: “Our business is different. Settlement timelines make real ROI tracking impossible.”
Another industry faced an almost identical problem — and solved it fifteen years ago. That industry is property and casualty insurance. The structural parallels are close enough that the solution maps almost directly onto PI marketing. And yet most firms have never looked outside their own vertical for the answer.
How the Insurance Industry Used to Think About Acquisition
In the early 2000s, P&C insurance companies measured marketing performance the same way most PI firms measure it today: cost per new policy. They tracked acquisition cost across every channel — direct mail, online ads, independent agents, comparison sites — and optimized for the cheapest number.
The problem was obvious in hindsight. A policyholder acquired for $180 through a comparison site might cancel after one renewal. A policyholder acquired for $420 through a local independent agent might stay for nine years, add auto and umbrella coverage, and generate $14,000 in lifetime premium revenue. The cheaper acquisition was the worse investment by every measure that actually mattered.
The industry couldn't see that because they were measuring the wrong thing at the wrong time horizon. Cost per acquisition showed how efficiently they were buying customers. It said nothing about whether those customers were worth buying.
The Shift: From Acquisition Cost to Revenue by Channel
Keep reading
Between roughly 2008 and 2013, the major P&C carriers — Progressive, GEICO, Allstate, State Farm — stopped optimizing for cost per new policy and started optimizing for what they called “lifetime customer value by acquisition channel.”
The reframe was simple: instead of “which channel gives us the cheapest new policy?” they asked “which channel produces policyholders who generate the most premium revenue over 36 months?” That single question changed everything about how they allocated budget.
Channels that looked expensive turned out to be the highest-ROI investments when measured against long-run revenue. Channels that looked cheap turned out to produce high-churn customers who cost more to service than they generated in premium.
- Comparison-site leadswere cheap to acquire but had the highest churn rates — these customers were price-shopping by nature and left at the next renewal.
- Agent-referred policyholderscost more upfront but had 3x higher retention rates, added more products, and generated significantly more lifetime revenue.
- Brand-direct inquiriesfrom advertising fell in the middle — moderate acquisition cost, moderate retention, moderate lifetime value.
Once carriers could see this, budget allocation shifted dramatically. They didn't abandon comparison sites entirely — but they stopped treating comparison-site cost-per-acquisition as the benchmark every other channel had to beat.
| P&C Insurance | Personal Injury | |
|---|---|---|
| Acquisition metric | Cost per new policy | Cost per signed case |
| Revenue lag | 12–36 months (renewals) | 6–18 months (settlements) |
| Value variance by source | 3–5x across channels | 5–20x across vendors |
| Cheap-source trap | High churn, low lifetime value | High attrition, low settlement value |
| Solution | Lifetime value by channel | Cost per settled case by vendor |
Why the Structural Economics Mirror Personal Injury
This is not a metaphor. The structural economics are genuinely parallel. Look at the five features that made insurance attribution hard:
- Long revenue lag.Insurers don't know a customer's true value for 24 to 36 months. PI firms don't know a signed case's true value for 6 to 18 months — until settlement. Same timing gap, different industry.
- Multiple simultaneous acquisition channels.Carriers managed 8 to 12 channels at once. PI firms spending $200K or more per month typically run 5 to 10 lead vendors alongside organic and referrals. The complexity is comparable.
- High and variable cost per acquisition.Insurance acquisition costs ranged from $150 to $600 by channel. PI firms see cost per signed case range from $1,500 to $8,000 or more by vendor and case type. Misallocation is expensive in both models.
- Vendor-reported metrics that obscure the truth. Comparison sites reported conversion rates that made their channel look efficient. PI lead vendors do the same — cost per lead, maybe cost per signed case, but never cost per settled case or revenue per dollar spent.
- Channel quality varies invisibly.A $200 lead that produces a $350,000 settlement is a fundamentally different asset than a $150 lead that produces a $40,000 policy-limits case or gets withdrawn six months in. Without settlement data, you cannot see the difference.
This is a structural match. The same forces that made cost-per-policy a misleading metric for insurers make cost-per-lead — and even cost-per-signed-case — a misleading metric for PI firms.
The Key Insight: Optimize for Revenue, Not Acquisition Cost
The insurance industry's breakthrough was not a technology innovation. It was a measurement innovation. They changed what they were optimizing for.
Before:Minimize cost per new policy across all channels.
After:Maximize premium revenue per marketing dollar over a 36-month attribution window.
That shift required two things. First, connect acquisition data to downstream revenue — linking the channel a customer came from to the premium that customer generated over time. Second, wait long enough to measure it. A 30-day window is useless. You need 12, 24, 36 months of revenue data to see which channels are actually producing value.
For PI firms, the translation is direct:
- Before:Minimize cost per lead. Maybe track cost per signed case. Grade vendors monthly on volume and conversion rate.
- After:Maximize settlement revenue per marketing dollar over a 12-to-18-month attribution window. Connect each lead source to the cases it produced, the settlements those cases generated, and the total revenue returned per dollar invested.
This is “lifetime customer value by acquisition channel” — adapted for a business where “lifetime” means the case lifecycle from lead to settlement.
What PI Firms Can Apply Directly
You do not need actuarial models. You need the same fundamental framework. Here is what that looks like in practice:
Track cost per case all the way to settlement.Not cost per lead. Not cost per signed case. Cost per settled case, with the actual settlement amount attached. This is the PI equivalent of lifetime customer value by channel. Without it, you are making allocation decisions the same way insurers did in 2005 — and they were wrong.
Extend your evaluation window.Grading vendors on 30- or 60-day performance means grading on lead volume and early conversion rates. That is like an insurer grading an agent channel on first-month policy count without checking renewal rates. Give attribution data 6 to 12 months to mature before making major reallocation calls.
Accept that your cheapest channel may be your worst investment.This is the hardest shift for directors who have spent years optimizing for cost per lead. The vendor with the lowest cost per lead might produce cases that settle for less, withdraw at higher rates, or drag on longer. The vendor with the highest cost per lead might be producing the best-value cases your firm signs all year.
Connect case management to marketing spend data. Carriers invested in linking policy administration systems to their attribution platforms. PI firms need to do the same — connect LeadDocket, Filevine, or your CMS to your vendor spend. Without that link, you are guessing, regardless of how many tabs are in the spreadsheet.
Report in revenue terms, not activity terms.Insurance executives stopped hearing “new policies per channel” and started hearing “premium revenue per marketing dollar by channel.” Your managing partner does not need lead counts. They need settlement revenue per dollar spent, by source.
Why “Our Business Is Unique” Is Only Half True
The objection that keeps PI firms stuck: “We're not insurance. Our cases are different. Every case is unique. Settlement amounts are unpredictable.” All of that is true at the individual case level. None of it is true at the portfolio level.
Insurers face the same objection internally. Every policyholder is different. Claims are unpredictable. Behavior varies by geography, demographics, and weather events. At the individual level, prediction is hard. At the portfolio level — thousands of customers by channel over 36 months — the patterns are clear, consistent, and actionable.
PI marketing works the same way. One case from Vendor B might settle for $15,000 and another for $1.2 million. But across 50 to 100 cases from that vendor over 12 months, average settlement value, withdrawal rate, time to resolution, and cost per settlement dollar all stabilize into patterns you can measure and act on.
Your business is unique in its specifics. Not in its structure. And it is the structure — long revenue cycles, multi-channel acquisition, high and variable costs — that determines which measurement approach works. The insurance industry proved the right approach is revenue attribution by channel over an extended time horizon. The solution is structurally identical for PI.
The Cost of Staying in 2005
Carriers that were slow to adopt revenue-by-channel attribution did not just miss an optimization opportunity. They actively misallocated capital for years. They poured budget into channels that produced cheap, low-value customers and starved the channels that produced expensive, high-value ones. First movers built a compounding advantage that late adopters never fully closed.
PI firms still optimizing for cost per lead — or cost per signed case without settlement data — are making the same structural error. They are allocating $200,000, $400,000, $600,000 per month across vendors without the one number that tells them which dollars are working: revenue by source, all the way to settlement.
The insurance industry solved this problem. The framework exists. The structural economics are the same. The only question is whether your firm adopts the solution now, or spends another year allocating budget the way Progressive did in 2005.
Related guide:For the partner-level conversation this analysis is designed to enable, see The Managing Partner's Guide to Marketing ROI — the metrics, the reports, and the budget conversations every PI leadership team should be having quarterly.