What is a good ROI for law firm marketing?

There is no transferable benchmark, because the answer is set by your fee share and average case value rather than by your marketing. Picture two firms running identical campaigns, one collecting $6,500 a case and the other $60,000: they will report returns an order of magnitude apart. Use your own break-even instead: if cost per signed case exceeds the fee you collect on that case type, the channel loses money, and every figure above that is a judgement about how much margin you want.

Should I use case value or attorney fees when calculating ROI?

Attorney fees, collected. The client's gross recovery is not your revenue. On a one third contingency it is three times your fee, so putting it in the numerator triples the revenue side of the calculation and inflates the return by even more than that. On hourly and flat fee work, use collected rather than billed, net of write-offs. Expected fee is acceptable for a leading indicator at signature provided you label it as expected and reconcile it against collections later.

How do I count a referral case if the client also clicked one of our ads?

Ask at intake and record the answer as the primary source, with the ad click stored separately as an assist. A referred client searching your firm name and clicking a brand ad is confirming a decision already made, not being acquired by that ad. Last click attribution will credit the ad every time, which is how brand campaigns end up looking like the best performer in the account.

How long should I wait before judging a channel's ROI?

Long enough to cover your own signature-to-resolution cycle, which you should measure from your own closed matters rather than assume from a published range. Track two numbers rather than waiting: signed ROI at signature using expected fee, which you can act on within weeks, and collected ROI once matters resolve, which is the real figure. Judging a channel on a single quarter of collections is judging spend you made long before that quarter.

Can Google Analytics calculate signed case ROI on its own?

No. GA4 can record the form submission, the call and the channel that delivered them, and that is where it stops, because the signature happens in your case management system weeks later and nothing tells GA4 about it. You get signed case ROI by exporting signed matters with their source, joining them to spend, and optionally sending the signed case back into the ad platforms as an offline conversion.

What if our case management system has no marketing source field?

Add one as a required picklist on the matter record and start collecting forward from today. Backfilling is possible by matching intake dates and phone numbers against your lead records, but the match rate falls quickly the further back you go, so treat historical figures as directional. Two clean quarters of forward data are worth more than a reconstructed year.

Part of: Marketing attribution

How to measure signed case acquisition ROI

By Mohammad Kashif, Chief Technology OfficerLast updated September 21, 2026

Signed case acquisition ROI is (fee revenue from signed cases minus channel spend) divided by channel spend. Use the fees your firm actually collects, not gross settlements, and credit each case to the source recorded at intake.

A worked example with invented figures, not a client result: $12,000 spent, 86 leads, 29 consultations, 8 signed cases at $6,500 in collected fees. That is $52,000 back on $12,000 out, a 3.3x return and $1,500 per signed case. Swap in your own five numbers and the arithmetic is identical.

What is the formula for signed case acquisition ROI?

Return on a marketing channel is fee revenue from the cases it produced, minus what the channel cost, divided by what the channel cost. Express it as a multiple (3.3x) or as a percentage (333%), but pick one and stay with it, because a firm quoting both in the same report is usually quoting the same number twice and calling it two results.

The numerator is where most calculations quietly inflate. Revenue means the fee your firm collects, not the client's recovery. On a one third contingency, a $19,500 settlement is $6,500 of fee and $13,000 that was never yours. Using the recovery instead of the fee triples the number in the numerator and, because the return is net of spend, inflates the result by even more than that. On hourly or flat fee work the equivalent trap is billed rather than collected: use collected, net of write-offs and courtesy discounts, or you are measuring optimism.

The denominator is everything you paid to keep the channel running during the period. If your spend is a retainer, the retainer is the spend: a Starter plan at $599 a month is $1,797 across a quarter, and a single-channel agency retainer at $4,000–$10,000 a month is $12,000 to $30,000 across the same quarter. Paid ads sit on top as media cost. What you include matters less than including the same things every time you run the calculation.

Media cost as the platforms actually billed it, not the budget you set.
Retainers, subscriptions and licences for the months in the period.
Content and asset production paid for in the period, whether or not it published in the period.
Call tracking numbers, form and scheduling tools, anything you would cancel if you switched the channel off.
Staff time only if you are willing to carry it across every channel including referrals, which most firms are not. Leaving it out consistently beats including it selectively.

What does the calculation look like step by step?

Take one channel and one quarter. The figures below are placeholders chosen so every step of the arithmetic is visible. They are not a benchmark and nothing here claims a firm should expect them.

Spend was $12,000. The channel produced 86 leads, which is $139.53 per lead. Intake booked and held 29 consultations from those 86 leads, a 33.7% booking rate. Eight of the 29 consultations signed, a 27.6% close rate, which is 9.3% of leads end to end. Average collected fee per signed case was $6,500, so fee revenue is 8 at $6,500, or $52,000. Subtract the $12,000 and divide by the $12,000: the return is 3.3x, or 333%. Cost per signed case is $12,000 divided by 8, or $1,500.

The single ROI figure is the least useful number in that paragraph. It tells you the channel paid, and nothing about what to do next. The intermediate rates tell you where to act. Hold spend and lead volume exactly where they are and move the close rate from 27.6% to 17.2%: five cases instead of eight, $32,500 instead of $52,000, a 1.7x return instead of 3.3x, and cost per signed case up from $1,500 to $2,400. Nothing changed upstream of the consultation. Intake moved the return by half.

That is the practical reason to compute this per channel rather than across the firm. A blended return hides the channel that produces cheap leads nobody signs, which is how a marketing budget gets spent badly while every report stays green.

Where does signed case ROI go wrong in a law firm?

The formula is arithmetic and the arithmetic is never the problem. Three data joins break it, and all three are structural rather than careless.

The first is the missing join between the two systems. Signed cases live in the case management system. Spend and leads live in the ad platforms and the analytics property. Nothing joins them on its own, so the firm ends up holding two reports that are each correct and cannot be multiplied together: in the example above, marketing knows it produced 86 leads, the practice knows it signed 8 cases, and nobody can say which 8 of the 86. Cost per lead gets quoted to the cent and cost per signed case gets estimated.

The second is referral contamination. A client who was referred by a former client will still search your firm name, click the brand ad sitting above your own listing, and call from that page. Last click hands that case to paid search. The channel that produced nothing is credited with the fee, and the referral relationship that actually produced it shows a zero. Where referrals carry a large share of the work, this distortion runs in one direction only: it argues for spending more on the channel that deserves it least.

The third is time. Contingency matters sign in one quarter and pay in another, and the gap between the two is long enough that your own closed-matter records are the only reliable guide to it. Fees collected this quarter came from spend you made in earlier periods, at budgets and on keywords you have since changed. Dividing this quarter's collections by this quarter's spend divides two unrelated populations and produces a number that moves for reasons that have nothing to do with marketing.

The tell for the missing join: you can state cost per lead exactly and cost per signed case only as a guess.
The tell for referral contamination: paid search posts a strong month whenever a good referral signs, and brand terms carry an implausible share of your signed cases.
The tell for the timing mismatch: return swings hard quarter to quarter while spend and lead volume sit flat.

How do you join case management data to marketing spend?

The fix is one required field and one identifier that survives both systems. Neither is technically hard and both fail for the same reason, which is that they have to be filled at intake by the person taking the call rather than reconstructed later by whoever runs the report.

Add a source field to the matter record as a required picklist, not free text. Free text produces "web", "Website", "google" and "Google search" as four separate channels, and merging them afterwards is guesswork. Keep the list short enough that intake picks the right option under time pressure: paid search, organic search, referral, directory, repeat client, other. Ask the caller directly, record what they say, and record the digital touch separately as an assist rather than overwriting the answer with it.

For the identifier, the normalised phone number usually works, because it exists in the lead record and in the matter record. Distinct tracked numbers per channel make the phone route reliable rather than inferred. Then reconcile monthly: export signed matters with their source and signed date, join them to spend by channel and month, and recompute. This is a spreadsheet job before it is a systems job, and running it by hand for two quarters teaches you which fields are actually trustworthy.

Then push the result back the other way. Ad platforms accept offline conversion imports, so a signed case can be sent back as the conversion the bidding optimises toward. Left alone, the platform optimises for form fills, and nothing in that signal separates a form fill that becomes a case from one that never will. GA4 is worth configuring and will not solve this by itself: it can record the form fill and the call, but it cannot know the case signed unless you feed the signed outcome back into it.

How long before the number actually means anything?

Cohort by the spend period, not by the calendar. The question is what the money you spent in Q1 eventually produced, so follow the Q1 spend forward through the cases it signed, rather than asking what arrived during Q1 from spend of unknown vintage.

Run two checkpoints and label them so they never get mixed. Signed ROI uses expected fee at signature and is available within weeks, which makes it the number you can actually steer with. Collected ROI uses the fee that landed in the account and is available when matters resolve, which makes it the number that is true. A firm reporting only signed ROI is overstating revenue by whatever share of expected fees it never collects; a firm reporting only collected ROI is steering by a rear-view mirror as deep as its own case cycle.

Then be honest about sample size, because this is where small firms overreact. In the worked example above, one additional signed case moves the return from 3.3x to 3.9x and one fewer takes it to 2.8x. At eight signed cases a quarter, the figure is not precise enough to justify rebuilding a budget on a single period. Judge channels on four quarters, or judge them on the intermediate rates, which run on a denominator of 86 leads rather than 8 cases and therefore settle down far sooner.

Signed case acquisition ROI worked end to end: one channel, one quarter, placeholder figures

Step	Figure	How it is derived
Channel spend	$12,000	Retainer plus media plus channel tools, three months
Leads attributed	86	Unique inbound contacts tagged to the channel
Cost per lead	$139.53	$12,000 divided by 86
Consultations held	29	33.7% of leads, the intake booking rate
Signed cases	8	27.6% of consultations held, 9.3% of leads
Average collected fee	$6,500	One third of a $19,500 recovery, the fee and not the recovery
Fee revenue	$52,000	8 signed cases at $6,500 each
Cost per signed case	$1,500	$12,000 divided by 8
Return	3.3x, or 333%	($52,000 minus $12,000) divided by $12,000

Common questions

What is a good ROI for law firm marketing?: There is no transferable benchmark, because the answer is set by your fee share and average case value rather than by your marketing. Picture two firms running identical campaigns, one collecting $6,500 a case and the other $60,000: they will report returns an order of magnitude apart. Use your own break-even instead: if cost per signed case exceeds the fee you collect on that case type, the channel loses money, and every figure above that is a judgement about how much margin you want.
Should I use case value or attorney fees when calculating ROI?: Attorney fees, collected. The client's gross recovery is not your revenue. On a one third contingency it is three times your fee, so putting it in the numerator triples the revenue side of the calculation and inflates the return by even more than that. On hourly and flat fee work, use collected rather than billed, net of write-offs. Expected fee is acceptable for a leading indicator at signature provided you label it as expected and reconcile it against collections later.
How do I count a referral case if the client also clicked one of our ads?: Ask at intake and record the answer as the primary source, with the ad click stored separately as an assist. A referred client searching your firm name and clicking a brand ad is confirming a decision already made, not being acquired by that ad. Last click attribution will credit the ad every time, which is how brand campaigns end up looking like the best performer in the account.
How long should I wait before judging a channel's ROI?: Long enough to cover your own signature-to-resolution cycle, which you should measure from your own closed matters rather than assume from a published range. Track two numbers rather than waiting: signed ROI at signature using expected fee, which you can act on within weeks, and collected ROI once matters resolve, which is the real figure. Judging a channel on a single quarter of collections is judging spend you made long before that quarter.
Can Google Analytics calculate signed case ROI on its own?: No. GA4 can record the form submission, the call and the channel that delivered them, and that is where it stops, because the signature happens in your case management system weeks later and nothing tells GA4 about it. You get signed case ROI by exporting signed matters with their source, joining them to spend, and optionally sending the signed case back into the ad platforms as an offline conversion.
What if our case management system has no marketing source field?: Add one as a required picklist on the matter record and start collecting forward from today. Backfilling is possible by matching intake dates and phone numbers against your lead records, but the match rate falls quickly the further back you go, so treat historical figures as directional. Two clean quarters of forward data are worth more than a reconstructed year.

Find out what your signed cases actually cost

We trace your last quarter of signed cases back to the channel that produced them, and show you the cost per case next to each one.

Get an attribution audit