Why Cost Per Lead Is Harder to Track Accurately Than Most PI Firms Realize

Cost per lead seems like the simplest metric in marketing. Take what you spent, divide by the number of leads you got. The math is easy. But for PI firms managing multiple vendors across different channels, calculating an accurate cost per lead is far more complicated than most people expect — and the errors in that calculation have real downstream consequences.

Here's what makes cost per lead tracking harder than it looks, and why those inaccuracies matter even though cost per lead isn't your most important metric.

The Definition Problem: What Counts as a Lead?

Before you can calculate cost per lead, you need a consistent definition of “lead.” That sounds obvious — but in practice, different vendors, different channels, and different internal teams are all using different definitions.

Consider the variation in a typical PI firm's vendor mix:

• Pay-per-call vendors typically count any call that meets a minimum duration (often 60 or 90 seconds) as a billable lead. That threshold is designed to filter out wrong numbers and immediate hang-ups — but a 90-second call can still be completely unqualified.
• Web form vendors often count any form submission as a lead, regardless of whether the contact information is valid or the prospect is actually in the right jurisdiction.
• LSA (Local Services Ads)counts “leads” as calls or messages routed through the platform — but you can dispute leads that don't meet Google's criteria, which means your effective cost per lead can change after the fact depending on how aggressively you manage disputes.
• Internal intake teams might define a lead as any prospect that expressed potential interest — which could be a much lower bar than what your vendors are billing you for.

When you add all of these together in a monthly summary report, you're comparing apples to oranges. The blended “average cost per lead” number is meaningless because the inputs aren't measuring the same thing.

The Duplicate Lead Problem

Duplicate leads are more common than most firms realize, and they inflate your lead count while artificially lowering your apparent cost per lead.

A common scenario: a prospect sees your TV ad, then searches for your firm by name, then clicks a Google search ad, then calls the number on your website. That one prospect may generate a billable event from your TV vendor (for the brand awareness they drove), a click charge from Google, and an inbound call recorded by your intake team. You've “received three leads” from one person who was already trying to reach you.

Without a deduplication process — matching leads across systems by phone number, email, or name — your lead count is inflated and your cost per lead is understated. And when the same person calls two different vendors who then both claim credit for the case you signed, your double-counted lead becomes a double-counted case.

The Billing Reconciliation Gap

Many PI firms run their marketing budgets on a mix of contracts, retainers, and per-lead billing. Reconciling what you were actually billed against what you actually received is harder than it sounds.

A pay-per-call vendor might send you a monthly invoice showing 87 leads at $225 each = $19,575. But when you check your call log, you show 94 inbound calls from that vendor's tracking numbers that month. Some of those calls weren't billed (short duration, filtered out), but some calls you received weren't in the vendor's invoice. Which number is right? What's your actual cost per lead for that vendor?

These discrepancies exist in nearly every vendor relationship and rarely get investigated carefully. The result is that most cost per lead calculations are based on invoice totals rather than verified delivery counts — which means they're only as accurate as the invoices.

Channel Attribution Affects Lead Counts

Your internal lead count can diverge from your vendor-reported lead count based on how your intake team attributes incoming leads.

If your intake system asks “how did you hear about us?” and records the answer, you'll sometimes see situations where the prospect says “TV” but the call came through a digital tracking number that your system attributes to Google. Which attribution is right? Neither is wrong — they're measuring different touchpoints in the same prospect's journey.

The practical problem: your internally-reported lead count from each source won't match what your vendors report, and both will diverge from what you were billed. When you divide total spend by internal lead count, you get one cost per lead. Divide by vendor-reported lead count and you get another. The difference can be 20 to 40% in either direction.

Timing Differences Distort Monthly Reports

Most PI firms track cost per lead on a monthly basis. But billing cycles and lead delivery don't align perfectly to calendar months, which creates apparent swings in performance that aren't real.

A vendor might bill net-30, meaning their January invoice arrives in early February and gets recorded as a February expense — even though the leads it covers were delivered in January. If January was a high-volume month and February was slow, your February cost per lead report shows high spend and low leads: terrible performance. But you're actually looking at January's leads and February's invoices. The comparison is meaningless.

Accrual-based accounting handles this properly, but many PI firms operate on a cash basis for expense tracking — which means their marketing performance reports have systematic timing distortions baked in.

Why These Inaccuracies Matter Even Though Cost Per Lead Isn't the Right Metric

At this point you might be thinking: “If cost per lead is already the wrong metric to optimize for, why does it matter that we can't track it accurately?”

It matters for a few reasons:

• Cost per case is calculated from cost per lead. If your cost per lead inputs are wrong, your cost per case calculations are wrong. Errors compound upward.
• Billing disputes start with accurate lead counts. If you're paying for leads you didn't receive — or receiving leads you weren't billed for — you can't manage that without accurate tracking.
• Trend analysis requires consistent definitions.If your definition of “lead” shifts over time — even slightly — your month-over-month comparisons are measuring definitional changes, not real performance changes.

Building a More Accurate Cost Per Lead Foundation

If you want to track cost per lead accurately enough to be useful, a few practices help significantly:

• Define “lead” once, in writing, for all purposes. Pick a definition — a qualified inbound contact that meets a minimum threshold of practice area fit and jurisdictional eligibility — and apply it consistently across all vendors and internal tracking.
• Reconcile invoices against delivery data monthly. Compare what the vendor billed against what your systems recorded. The discrepancies will tell you where your tracking has gaps.
• Deduplicate before counting. Match leads across channels by phone number and email before adding them to your totals. One person is one lead, regardless of how many channels they touched.
• Use accrual-period matching. Record expenses in the month the leads were delivered, not the month the invoice arrived.

Cost per lead will never be your north star metric — cost per case is what matters for vendor decisions. But accurate cost per lead data is the foundation that cost per case calculations are built on. If the foundation is wrong, everything above it is unreliable too.