What Data Do You Need Before You Can Start Tracking Cost Per Case?

Here's the wall most PI firms hit when they first try to track cost per case: the spend data lives in QuickBooks, the case data lives in LeadDocket, and lead source attribution lives in a spreadsheet — if it lives anywhere at all. When you try to connect them, the numbers contradict each other. You can't tell which ones to trust.

Decisions made on bad numbers are worse than no numbers at all. That's the real risk of rushing into cost-per-case tracking before your data foundation is solid.

This is your pre-flight checklist: exactly what data you need, where it typically lives, and what to do when it's messy — because it will be.

The Three Categories of Data You Need

Cost per case connects two inputs: what you spent and what you got. Simple in theory. Hard in practice — because PI firms typically have spend data in one place, case data in another, and the connection between them nowhere.

Here's what you need across each category:

1. Marketing Spend Data

This is your total investment by source. Not an estimate — actual spend. For each lead vendor and marketing channel, you need:

• Monthly spend by vendor— what you actually paid, not what was quoted or budgeted
• Start and end dates for each vendor relationship— to align spend with the right time periods
• How each vendor bills— flat monthly retainer, per-lead pricing, performance-based, or hybrid
• Any credits, refunds, or adjustments— these shift your true cost per case and matter at scale

Most firms have this data. It's just scattered: invoices buried in email, entries in QuickBooks, totals on a marketing spreadsheet, and some figures only accessible inside the vendor's portal. Step one is consolidating everything into a single ledger — even a rough one — before connecting it to case outcomes.

2. Lead and Case Data

This is the output side of cost per case. For each lead that enters your pipeline, you need:

• Lead source— which vendor or channel sent this lead
• Lead received date— when it entered your system
• Lead disposition— signed case, rejection, or withdrawal
• Signed case date— when the case was signed, not when the lead arrived
• Case type or severity— if your firm tracks severity, this enriches analysis significantly

This data should live in your case management software — LeadDocket, Filevine, Clio, MyCase. The question is whether it's structured. If your intake team records lead source as free text (“TV ad” vs. “TV” vs. “television”), you have a data cleaning problem before you have a tracking problem.

3. Attribution Data

Attribution bridges spend and cases. It answers a single question: which spend produced which cases? For most PI firms, this is the weakest link in the chain.

Attribution data can come from several sources:

• Call tracking(CallRail or similar) — ties inbound calls to their marketing source
• Form tracking— web forms with UTM parameters or source fields that capture lead origin
• Intake questions— specialists asking “How did you hear about us?” and recording the answer in a structured field
• Vendor-assigned lead IDs— some vendors tag each lead so you can trace it back to their specific campaign

Perfect attribution isn't the goal. Consistent attribution is. The same methodology applied across all vendors is what lets you compare them fairly. A firm that asks “How did you hear about us?” and records the answer in a dropdown beats a firm using call tracking for some sources and intake questions for others — because at least the methodology is uniform.

How Much Historical Data Do You Need?

One month of clean, connected data is enough to start. But you need at least three months before patterns become meaningful — and six months before you can make reliable vendor decisions based on trends rather than point-in-time snapshots.

PI firms face a specific complication here: the 6 to 18-month settlement lag means your full financial picture — spend connected to settlement revenue — won't stabilize for 6 to 9 months after you start tracking. But your cost-per-signed-case data becomes meaningful much sooner. Start there.

If you have 12 months of historical spend and a case management system with structured lead source and disposition data, you can build a retroactive cost-per-case baseline for the past year. That's a significant head start — you're not beginning from zero.

What to Do When Your Data Is Messy

Messy data is the norm, not the exception. Here's how to work through the four most common problems:

Inconsistent Lead Source Labels

“TV ad,” “TV,” and “television” in your case management system are three separate sources as far as any analysis tool is concerned. Consolidate before you analyze. Build a source mapping document that standardizes vendor names across all systems, and lock intake specialists to a dropdown going forward — no more free-text entry.

Missing Spend Records

No exact figures for older periods? Use estimates from bank records, accounting software exports, or vendor invoices. Within 10% is close enough to build a baseline. Flag where estimates were used so you can refine them as cleaner records surface.

No Lead Source Data for Older Cases

For cases that predate your tracking improvements, attribute them in aggregate. If 60% of your historical spend went to Vendor A and you signed 100 cases in that period, proportional attribution gives you a working estimate. It's imprecise — but it produces a usable baseline while your new practices accumulate clean data.

Cases Without Signed Dates

If your system captures case status but not the date the case was signed, use case creation date as a proxy. Document the substitution. As you tighten intake, start capturing signed dates explicitly — they're essential for any time-based cost-per-case analysis.

The Data Minimum Viable Product

You don't need perfect data. You need a minimum viable dataset:

• Total spend per vendor for the past 6 to 12 months (estimated is fine)
• Total leads received per vendor for the same period
• Total signed cases attributable to each vendor for the same period
• A standardized vendor naming convention going forward

With those four inputs, you can calculate a rough cost per case by vendor today. It won't be clean. It will tell you things you didn't know. And it gives you a baseline to sharpen as your tracking improves.

An intake manager at a 25-attorney PI firm described it this way: “We thought we needed six months to clean everything up before we could start. We started with what we had — imperfect data — and connecting it to cost per case immediately revealed where the gaps were. Fixing the data became a business priority once we could see what the clean version would show us.”

A Pre-Launch Data Checklist

Run through this before you start tracking cost per case:

• Historical spend by vendor: 6–12 months minimum, actual or estimated
• Standardized vendor/source names across all systems
• Lead disposition data in a structured field — not free text — in your case management system
• Attribution mechanism in place for all active lead sources (call tracking, UTMs, or intake question with a structured field)
• Signed case date captured in your CMS (or a reliable proxy)
• A defined cost-per-case target — what does “good” actually look like for your firm?

Check all six and you can start tracking with confidence. Check four or five and you can start with appropriate caveats about your data quality. Check two or three and fix the foundation first — wrong conclusions from bad inputs are more costly than the delay.

Ready to Connect Your Data?

RevenueScale is built for the data PI firms actually have — messy spreadsheets, gap-filled histories, partially structured case management exports. That's a real-world starting point, not a disqualifier.

Book a demo and we'll review your specific data situation: what's ready to use now, what needs cleanup, and what your cost-per-case tracking can realistically show within 30 days.