Legal Spend
What percentage of invoices should be rejected or adjusted?
Updated July 2026
There is no universal right percentage to target, and setting one as a goal usually distorts the program. A rejection or adjustment rate is only meaningful read against its trend and its cause. A single number tells you nothing on its own, because a low rate can mean firms are billing clean or that no one is really checking. Watch the direction and the reasons, not a target figure.
Why is there no correct target percentage?
Because the rate depends on your firms, your guidelines, and how enforcement works, so a number that is healthy for one program is a warning sign for another. A rejection rate is an output of many inputs, not a dial you set. Chasing a specific figure pushes people to hit the number rather than fix what the number reflects, which is why trend and cause are the useful frame.
A borrowed industry percentage is a prior at best, not a target. It tells you roughly what other programs see, under conditions you cannot verify, with firms and guidelines that are not yours. Treating it as a goal means optimizing for a number that was never a measure of whether your own spend is under control.
Why can a low rejection rate be good news or bad news?
Because the same low number has two opposite explanations. It can mean your firms bill cleanly to well-enforced guidelines, so there is genuinely little to adjust. Or it can mean no one is really checking, and violations are being paid because invoices are approved on trust. The rate alone cannot tell you which. You have to know whether real enforcement is happening behind the number.
| What you observe | Healthy explanation | Warning explanation |
|---|---|---|
| Low rejection rate | Firms bill clean against enforced guidelines, little to adjust | Invoices approved on trust, violations paid because no one checks |
| High rejection rate | Enforcement is catching real leakage that was there all along | Guidelines are unclear or firms are not managed, and disputes pile up |
| Falling rejection rate over time | Firms have adjusted to consistent enforcement, behavior improved | Enforcement quietly lapsed and problems stopped being caught |
What should legal ops actually measure instead?
Measure the trend and the cause. Track whether the rate is rising or falling and why, break adjustments down by violation type and by firm, and watch whether the same firms generate the same problems after being flagged. A rate moving in the right direction for a reason you understand is worth more than any single quarter's figure measured against an outside benchmark.
- Direction over time, not a single quarter, since the trend shows whether enforcement is working.
- Adjustments by violation type, so you know whether block billing, rate creep, or staffing drives the number.
- Adjustments by firm, so persistent offenders are visible and can be managed or reassigned.
- Whether flagged firms improve after being raised, which is the real test of whether enforcement changes behavior.
How does front-loaded enforcement change what the number means?
It moves most corrections upstream, so the rejection rate stops being the main scoreboard. When guidelines are enforced before invoices are even submitted, firms bill correctly the first time and there is less to reject at payment. A falling rejection rate then reflects better firm behavior, not weaker checking. That only holds if you can prove enforcement is actually running, which is why cause always outranks the raw percentage.
- Encode guidelines so firms see the rules before they bill, and correct their own entries first.
- Screen every submitted invoice line automatically, so nothing depends on which invoices got sampled.
- Read the rejection rate as one signal among several, alongside spend per matter and firm performance.
- Confirm enforcement is live before trusting a low number, since a low rate with no checking is the trap.
The goal is not to hit a target rejection percentage. It is to make the number a byproduct of a program where firms bill to enforced guidelines, corrections happen before payment where possible, and the trend and its causes are visible enough that a low rate can be trusted rather than assumed.
Common questions
Isn't there an industry-average rejection rate we can benchmark against?
Any figure you find should be treated as a rough prior, not a target, and honestly it is hard to compare programs meaningfully because the inputs differ so much. What a program rejects depends on its firms, how detailed its guidelines are, whether enforcement runs before or after invoices are submitted, and how adjustments are even counted. Two programs reporting the same rate can be in completely different shape, and one reporting a lower rate can be worse off if the low number comes from nobody checking. Benchmarking a rejection rate against an outside average tells you almost nothing actionable about whether your own spend is controlled. Spend your attention on your own trend and the causes behind it, which are things you can verify and act on, rather than on matching a number whose conditions you cannot see.
If we set a target rejection rate, what goes wrong?
You optimize for the number instead of the thing the number was supposed to reflect, and both directions of the target distort behavior. Set the target high and reviewers feel pressure to reject entries to hit it, generating friction with firms over marginal calls. Set it low and there is pressure to wave invoices through so the rate stays down, which is the same as not enforcing at all. Either way, the metric stops measuring the health of your billing and starts driving behavior toward the metric. Rejection rate is a diagnostic output, useful for spotting a trend or a problem firm, not a goal to manage toward. The goal is firms billing correctly to enforced guidelines. If you get that right, whatever rejection rate results is simply what it is, and it is trustworthy because you know what produced it.
How do we tell whether a low rejection rate is real compliance or missed enforcement?
Prove the enforcement is actually running, then read the causes behind the number. A trustworthy low rate sits on top of a process where every invoice line is screened against the guidelines before payment, not a fraction sampled by a busy reviewer. Check whether the screening is happening on all invoices or only some, whether the same firms once generated violations that have since dropped off after being flagged, and whether spend per matter and firm performance are moving in the same healthy direction. If enforcement is verifiably live and firms have visibly adjusted, a low rate is real compliance. If no one can say for sure that every invoice is being checked, a low rate is exactly what you would expect from invoices approved on trust, and it is the more dangerous of the two situations because it looks like success.
Billing violations to watch for→What matters more than the rejection rate for controlling spend?
The bigger levers sit upstream and downstream of the invoice. Upstream, clear and enforced guidelines mean firms bill correctly the first time, so there is less to reject at all. Downstream, the largest cost drivers are how matters are staffed, how long they run, and whether the right cases resolve early, none of which a rejection rate captures. A program can have an immaculate rejection rate and still overspend badly if its matters are mis-staffed or settle late. Use the rejection rate as one signal that your billing enforcement is working, then put real weight on firm performance measurement, matter budgeting, and early resolution, which is where total cost per outcome is actually decided. The invoice is where you stop leakage; the matter is where you win or lose the money.
How to reduce outside counsel spend→CaseGlide is the litigation intelligence platform for Fortune 500 legal departments and insurance claims organizations. It structures live litigation data from defense counsel into executive decisions: reducing defense spend, settling the right cases sooner, and shrinking litigated claim volume.
Next step · See it on your docket
See what your litigation portfolio is telling you
A 30 minute walkthrough on your own docket. No slides, no committee.