At this spend, a wrong ranking costs more than the reporting ever would.

A firm spending heavily across television, search, social and lead vendors is making the same allocation decision every month, usually on cost per signed case.

If that metric is wrong by a factor of two for one channel — and in this practice area it routinely is — the error repeats every month until somebody measures fees instead.

The difference is not marginal: fee distributions in personal injury are dominated by a small number of cases, and which channel produces them is not correlated with which produces volume.

Measuring fees by intake source across several years is the only way to see it, and at this budget the measurement pays for itself many times over.

The report gives the median beside the total so a source carried by one exceptional case is visible as exactly that.

A firm screening thousands of enquiries a month is paying intake staff, technology and management to say no most of the time.

That cost is entirely absent from cost per signed case, which means the cheapest lead source can be the most expensive one the firm buys.

Exporting intakes including rejections puts a real cost per signed case beside the fee revenue, which is the comparison that changes vendor decisions.

It also identifies sources whose rejection reasons cluster — wrong jurisdiction, outside statute, no injury — which is usually fixable at the source rather than by cutting the channel.

Fixing it is cheaper than replacing the channel, and neither option is visible without the data.

A case the firm refers to another for a share of the fee produces revenue with almost no cost of work, on a timeline the firm does not control.

Averaged with cases worked in-house it flatters whichever source produced it and tells you nothing about capacity or margin.

Splitting them gives two honest rankings, and most firms find the sources feeding each are different in a way that suggests a deliberate strategy rather than an accident.

Where the export flags the referral-out, this is automatic; where it does not, the fee pattern usually identifies them.

The distinction also matters for forecasting, because referred-out fees arrive on somebody else's schedule.

Fees arriving this quarter came from cases signed one to three years ago, so a report grouped by resolution date ranks marketing that is no longer running.

Grouping by intake date instead puts each cohort of signed cases against the spend that produced it, which is the only version a media buyer can act on.

It does mean the most recent cohorts are incomplete, and the report says so rather than showing a partial year as though it were final.

Stating that clearly is what keeps the report usable: an incomplete cohort presented as complete is how a good channel gets cut in its first year.