Why Don't My Marketing Reports Match My Case Management Numbers?

Almost every firm that looks closely finds this. Google Ads reports 60 conversions. The CRM shows 41 leads. Case management shows 12 signed cases. Three systems, three numbers, no obvious relationship between them.

The first thing worth saying is that this is normal, and the second is that normal does not mean fine.

The systems are counting different events

A marketing platform counts a conversion at the moment someone does something measurable — submits a form, clicks a call button, stays on a page long enough to trigger a goal. That event happens the instant the person acts.

Your case management system counts a case at the moment someone signs. That happens days or weeks later, after intake has spoken to them, after conflicts have been checked, sometimes after a medical record has been reviewed.

Between those two moments sits your entire intake process, plus a filter: most people who contact a law firm do not have a case that firm will take.

So the two systems are not measuring the same thing at different times. They are measuring genuinely different things. Expecting them to match is like expecting the number of people who walked into a shop to match the number who bought something.

Where the difference actually accumulates

The useful question is not why don’t these match but where does the difference sit, and how much at each point.

Duplicate counting. The same person fills in a form, then calls, then clicks an ad again a week later. To the platform that may be three conversions. To your firm it is one human being.

Non-conversions counted as conversions. Wrong-number calls, vendors, job applicants, people asking for a lawyer in a practice area you do not handle. Most platform conversion tracking cannot tell these apart from real inquiries.

Contacts that never reached anyone. After-hours calls with no voicemail, forms that failed to route, chats that timed out. These are invisible in the CRM but often counted by the platform, and they are the most expensive category, because they were real cases.

Unqualified inquiries. Real people with real problems that your firm would never take. These appear as leads and correctly never appear as cases.

Lost cases. Qualified people who did not sign, either with you or with anyone.

Attribution failures. Cases that came from a channel but were recorded as coming from nowhere, usually because the source field was blank on intake.

That last one is the quiet killer, and the one most firms underestimate.

The “unknown source” problem

If a meaningful share of your signed cases have no recorded source, every channel comparison you make is built on partial data. And it is rarely a random sample — unknown-source cases skew toward phone calls, walk-ins, and referrals, which means the channels that generate phone calls get systematically under-credited relative to the ones that generate form fills.

Firms in that position often conclude that a channel is underperforming and cut it, when what actually happened is that the channel produced cases nobody recorded properly. The decision looks data-driven. It is data-shaped, which is not the same thing.

What a good explanation sounds like

You do not need the numbers to match. You need to be able to account for the difference.

A firm that can say “the platform reported 60, about 12 of those were duplicates or wrong numbers, 41 became real lead records, 18 of those were cases we would take, and we signed 12” understands its own funnel. Every number in that sentence can be questioned, but the sentence exists.

A firm that can only say “the platform says 60 and we signed 12” has two numbers and no relationship between them. It cannot tell whether the problem is lead quality, intake, or measurement — and those three have completely different fixes.

The goal is not to make the systems agree. It is to be able to explain the distance between them.

Why nobody fixes it

Because it is not one person’s job. Marketing owns the platforms. Intake owns the CRM. The case management system belongs to the practice. Each group’s numbers are correct within their own system, and each can defend them. The gap only becomes visible when someone looks across all three at once, and in most firms nobody is asked to.

It is also genuinely tedious. Reconciling a month by hand means pulling exports from three or four systems and matching records that share no common identifier. Firms do it once, find it painful, and do not repeat it — so the picture goes stale immediately and the next reconciliation starts from scratch.

What to do first

Before anything else, find out how many of your signed cases have a recorded source. If that figure is low, fix it before analyzing anything, because every other analysis inherits the problem.

Then check whether every inbound contact is landing somewhere. Compare your call tracking total against your CRM lead count for the same month. The difference is contacts that reached your firm and vanished, and it is usually larger than firms expect.

Those two checks cost an afternoon and tell you whether your reporting problem is a measurement problem or a performance problem. Most firms assume the second and discover the first.