Measuring Return on Investment From Legal Technology
Law firms spend money on technology in the same way they often make other decisions: on the strength of a persuasive demonstration and a general sense that the tool will save time. Months later, nobody can say with confidence whether it did. Subscriptions renew by default, adoption varies from one fee earner to the next, and the original business case, if one existed, sits unread. Measuring legal technology ROI is the discipline that closes this gap, and it is considerably harder in a professional practice than in a manufacturing or retail business, because the principal input is the time of qualified people and the principal output is advice, which is not easily counted.
The difficulty is real but it is not a reason to avoid measurement. A firm that does not attempt to measure return is making an assumption that every software purchase pays for itself, and that assumption is frequently wrong. Some tools deliver a great deal more than they cost; others are used by a fraction of the people for whom they were bought; others merely relocate work from one person to another. A firm that cannot distinguish between these outcomes will not improve its next decision.
This article offers a practical method for a small or mid-sized firm. It explains what return means in a legal practice, how to establish a baseline, how to account honestly for cost, how to model benefits without inflating them, and how to review results after the tool has been in use. Worked figures that appear below are illustrative arithmetic only and are not market data or a statement of what any product costs or saves.
What return means in a law firm
Return on investment, in its simplest form, compares the net benefit of an investment with its cost. In a legal practice the difficulty lies in defining benefit, because the same hour saved may produce very different outcomes depending on the firm's business model.
Time recovered is not automatically money earned
A fee earner who saves an hour of administrative effort has not earned anything unless that hour is converted into something of value. It might be converted into an additional billable hour, into more careful work on an existing matter, into business development, into a shorter working day, or into nothing at all if the hour is simply absorbed by the slack in the day. The firm should state, for each tool, which conversion it expects and whether the evidence supports it.
For a firm that bills by the hour and has more demand than capacity, recovered time converts readily into billable work. For a firm that works on fixed fees, recovered time improves margin on each matter, because the fee is unchanged while the cost of delivery falls. For a firm that is short of work, recovered time has little direct financial value, but a tool may still be justified on grounds of quality or risk.
The four categories of benefit
Capacity and efficiency. Work that took longer now takes less time, or is performed by a less senior person at appropriate quality.
Revenue. The firm bills more, collects faster or reduces write-offs because of better time capture, billing discipline or client communication.
Risk reduction. The firm is less likely to miss a deadline, breach confidentiality, mishandle client money or lose a document. The benefit is the avoided cost multiplied by the reduced likelihood, which is difficult to quantify but cannot be ignored.
Quality and client experience. Clients receive faster responses, clearer information and a more professional service, which affects retention and referral. These benefits are real but the hardest to attribute to a single tool.
A balanced assessment of legal technology ROI considers all four, but assigns monetary value only to those that can be defended and treats the remainder as qualitative factors in the decision.
Establish a baseline before buying
The most common reason firms cannot measure return is that they did not measure the starting position. Without a baseline, any improvement is anecdote.
Choose a small number of measures
Resist the temptation to track everything. For each tool, select no more than three or four measures that relate directly to the problem it is meant to solve. The following are examples that a firm can usually extract from its own records.
The time between instruction and the opening of a matter, including conflict checks and engagement letter.
The proportion of fee earner time recorded against matters, compared with time worked.
The interval between completion of work and the issue of the invoice.
The average age of unpaid invoices and the volume written off.
The time taken to locate a given document or the latest version of a document.
The number of client enquiries about the status of a matter, and the time taken to answer them.
The number of near-misses in deadline management recorded in the preceding period.
Measure for long enough to be representative
A baseline taken over a single unusual week will mislead. Where the firm's work is seasonal or irregular, the baseline should cover a period long enough to include ordinary variation. If historical data are unavailable, a short period of deliberate sampling, in which a few fee earners record how long defined tasks take, is better than no baseline.
Record the baseline in writing
The baseline should be a short document, dated and held with the business case. It should state what was measured, how, over what period and by whom. Later comparisons must use the same method or they will not be comparable.
Account honestly for total cost
The subscription or licence fee is rarely the whole cost. A realistic assessment of legal technology ROI counts every cost that the firm would not have incurred without the tool.
Cost category | What to include |
|---|---|
Recurring charges | Subscription or licence fee, any add-on modules, storage charges, and any charges for additional features the firm will in practice need |
Implementation | Configuration, data migration, clean-up of existing records, integration with existing systems, and any paid onboarding |
Training | Time of every user spent in training, valued at an internal cost rate, together with time spent by whoever delivers the training |
Transition | Period of parallel running, reduced productivity during the learning period, and time spent resolving early problems |
Administration | Ongoing management of users, permissions, templates and workflows, normally the time of a named person |
Hardware and connectivity | Any devices, scanners or network improvements required to use the tool properly |
Exit | The cost of extracting data and moving to another system if the tool is later abandoned |
Internal time should be valued at a cost rate rather than a billing rate. The cost rate reflects what the firm actually pays for the person's time, including overhead; the billing rate reflects what the firm hopes to earn. Using the billing rate for training time overstates the cost; using it for recovered time overstates the benefit. The firm should be consistent and should state which it has used.
The pricing model is itself a part of the analysis. A tool charged per user increases in cost as the firm grows or as occasional users are added, whereas a flat price per firm does not. A firm should model cost at its present headcount and at a plausible future headcount, rather than at the headcount of the day on which the quotation was issued.
Modelling benefits without inflating them
Benefit models tend to be optimistic, because the people who prepare them are enthusiastic about the tool. Three habits keep the model honest.
Use the baseline and the actual task, not the vendor's claim
A claim that a tool saves a stated proportion of time on a task should be treated as a hypothesis to be tested on the firm's own work. The firm should run a short pilot with a few users on real matters and measure the result against the baseline. Evidence from the firm's own practice is more reliable than marketing assertions, and the firm should give no weight to figures that cannot be traced to a credible source.
Apply an adoption factor
A tool delivers benefit only to the extent that it is used. If half of the fee earners use it consistently, the aggregate benefit is closer to half of the theoretical maximum. The model should apply an explicit adoption factor, set conservatively, and the firm should revise it with real usage data after launch.
Apply a conversion factor to recovered time
As discussed above, an hour recovered is not necessarily an hour sold. The model should state what proportion of recovered time the firm expects to convert into billable or otherwise valuable work, and the reasoning. Where the answer is uncertain, a conservative assumption is preferable.
A worked example, expressed as illustrative arithmetic
The following example is illustrative arithmetic only. The figures are invented to demonstrate method; they are not market data, they do not describe any product, and they should not be used as a benchmark. A firm should substitute its own figures.
Suppose a small firm considers a system that improves time capture and billing. Assume the firm has five fee earners. Assume that, from its baseline, the firm finds that each fee earner fails to record a modest amount of billable time each week, and that the firm wishes to estimate the value of recovering a fraction of it.
Assume each fee earner fails to record 2 hours a week that would have been billable.
Assume the firm expects the tool to recover half of that, that is 1 hour a week per fee earner.
Apply an adoption factor of 80 per cent, because the firm expects that not every fee earner will use the tool consistently. This gives 0.8 hours a week per fee earner.
Apply a conversion factor of 75 per cent, because not all recorded time is ultimately billed at full rate after review. This gives 0.6 hours a week per fee earner.
Across five fee earners, the recovered billable time is 3 hours a week. Over 46 working weeks, this is 138 hours.
If the average realised rate is a stated amount per hour, the annual benefit is 138 multiplied by that rate. The firm should use its own realised rate, that is the amount actually collected per hour, not its standard rate.
The firm then sets this against the total cost for the year, comprising the recurring charge, a proportion of implementation, training and administration costs, and the transition cost. If the benefit exceeds the cost by a comfortable margin under conservative assumptions, the case is sound. If the case depends on every assumption being met in full, it is fragile. The firm should repeat the calculation with the adoption factor and the conversion factor reduced to test how sensitive the result is.
The calculation can be extended with benefits for faster invoicing, which accelerates cash collection and may reduce write-offs. Because cash timing and write-offs vary by firm, the firm should calculate them from its own ledger rather than from any general assumption.
Valuing risk reduction and quality
Risk benefits are the most difficult to value and the easiest to dismiss, yet for some tools they are the principal justification. A system that improves deadline management, conflict checking, document control or client money records reduces the likelihood of events that are very costly if they occur.
The firm should resist the temptation to fabricate a probability. A better approach is to identify the specific failure the tool is designed to prevent, to describe what it would cost the firm if it occurred, including professional indemnity excess, loss of the client, management time and reputational damage, and to ask whether the control is proportionate to that cost. The firm may also take into account any feedback from its insurer or regulator on the importance of the control.
Quality and client experience benefits should be recorded as qualitative factors unless the firm has data. Faster responses to client enquiries, a more consistent engagement process and a clear portal for sharing documents are real advantages, and they may be evidenced by client feedback and retention. They should be reported as such, without attaching a number the firm cannot defend.
Comparing options and the cost of doing nothing
ROI is a comparative exercise. The relevant question is rarely whether a tool has a positive return in isolation but whether it has a better return than the alternatives, including the alternative of continuing as at present.
The status quo. The cost of doing nothing is the continuing cost of the current problem, which the baseline should quantify as far as possible.
A simpler tool. A cheaper or narrower tool may capture most of the benefit. Features that are never used do not contribute to return.
A process change. Some problems are caused by process, not by lack of software, and a change in procedure may deliver the benefit at negligible cost. Technology applied to a poor process will often reproduce the poor process more quickly.
Consolidation. A single platform that replaces several subscriptions may deliver a better return than several specialist tools, because of lower integration cost and fewer logins, though the firm should confirm that the consolidated tool is adequate for each function.
The firm should also consider switching cost. A tool that holds the firm's data in a format that cannot readily be exported creates a long-term dependency, which is a cost even if it does not appear in any invoice. The firm should ask in advance how data would be exported and in what format.
Reviewing results after implementation
The business case is a forecast; the review is the test. A firm that does not review will repeat the same mistakes.
Set review points in advance
Agree at the outset when the review will occur, commonly at three months for adoption and at twelve months for financial return, and who is accountable for it. Record the measures that will be reviewed, taken from the baseline.
Measure adoption separately from outcome
If the outcome is disappointing, the first question is whether the tool was used. Usage data, where the system provides it, shows which users are active and which features are in use. A poor result with low adoption is a change management problem; a poor result with high adoption is a problem with the tool or with the original assumption.
Collect user feedback
Short structured conversations with users reveal friction that data do not, such as a required field that is routinely bypassed or a template that nobody trusts. Fix these promptly; they are often the difference between adoption and abandonment.
Decide: continue, adjust or exit
At the review, the firm should make an explicit decision. Continue as is, adjust configuration or training, or exit and reallocate the budget. Exit should be treated as a legitimate and sometimes wise outcome, and the firm should resist the sunk cost reasoning that keeps a poor tool in place because of the money already spent.
Common errors in measuring legal technology ROI
Counting gross time saved without netting the time spent on new tasks the tool introduces, such as data entry, administration or checking.
Valuing recovered time at the standard billing rate when the realised rate is lower.
Ignoring training, transition and administration costs.
Assuming full adoption on day one.
Attributing to one tool improvements that came from a concurrent change in staffing, process or workload.
Relying on figures from vendor marketing instead of the firm's own pilot data.
Failing to record a baseline, so that no comparison is possible.
Treating measurement as a one-off exercise at purchase, rather than a periodic discipline.
Practical next steps
For each tool under consideration, write a short statement of the problem it is meant to solve and choose three or four measures that relate directly to it.
Record a dated baseline from the firm's own records or from a short period of sampling.
Prepare a full cost schedule covering recurring charges, implementation, training, transition, administration and exit.
Run a small pilot with a few users on real matters and compare the results with the baseline.
Build the benefit model with explicit adoption and conversion factors, and test its sensitivity by repeating it with more cautious assumptions.
Appoint an owner for the investment and fix review dates at the outset.
Review usage and outcome at the agreed points and decide expressly whether to continue, adjust or exit.
Audit existing subscriptions once a year and remove those that cannot justify their cost.
How practice management software supports measurement
A platform that holds matters, time, documents, billing and client communication in one place makes the baseline and the review far easier, because the data for turnaround, billing lag, utilisation and aged debt are drawn from a single record. Legalytech provides this kind of integrated practice management environment for firms of different sizes, and its firm-wide pricing makes it simpler to model cost as headcount changes, but the quality of the measurement will always depend on the discipline with which the firm defines its baseline and reviews its results.
This article is provided for general information only and does not constitute legal advice. Practitioners should consult the rules of their own jurisdiction and regulator.
Comments