Enterprise Legal Management and the Spend Lever it Can’t Reach

A General Counsel opens the year-end report from the department's enterprise legal management platform. Outside counsel spend came in flat against last year. The audit rules recovered a few percentage points on invoices, block billing is down, and every matter carries a code and an owner. On the evidence in front of her, the platform did its job.

The flat number has two possible causes, and the report cannot tell them apart. Either the department paid less for the same volume of work, or it sent less work out in the first place. The first is enforcement, and it has a ceiling. The second is a change in what the department can absorb on its own, and it compounds every quarter. Enterprise legal management measures both and can produce only one.

That gap is getting expensive. The CLOC 2026 State of the Industry Report found only 37% of legal departments expect an increase in outside counsel spend, down sharply from 58% the year before. Hiring and outside counsel have both worked as release valves for a generation, and both are tightening at once. Meanwhile, in RSGI's 2026 research with Harvey customers, over a third of in-house teams report insourcing more work or instructing fewer firms. This piece covers where that lever sits, why enterprise legal management cannot reach it, and what changes in your reporting when work stops leaving the building.

What is Enterprise Legal Management and Who Owns it Inside a Company?

Enterprise legal management (ELM) is the discipline and software category corporate legal departments use to run legal work as an operation. An ELM platform brings matter tracking, outside counsel spend, e-billing, intake, and reporting into one place, so a legal department can see what work exists, what it costs, and who handles it.

In-house legal operations owns ELM inside most organizations. The General Counsel sponsors the program and answers for the numbers it produces. Finance and procurement are the constant counterparties, since spend data flows into enterprise reporting and the purchase itself runs through procurement.

The component list is stable across products. Matter management holds the inventory of active work. Spend management and e-billing check what firms submit against the billing guidelines your organization published. Outside counsel and provider management holds approved rates, panel assignments, and firm scorecards. Intake routes business requests to an owner. Reporting turns all of it into numbers a General Counsel can carry into a budget conversation.

Contract functionality appears in some platforms and not others, and the depth ranges enough to be worth asking about directly. Every one of the functions mentioned above operates on work that already exists. That constraint is the subject of the rest of this piece.

The Outside Counsel Lever Enterprise Legal Management Cannot Reach

Two ways exist to spend less on outside counsel. You can pay less for the work you send, or you can send less work. ELM is built almost entirely around the first.

Billing guidelines, automated invoice review, rate cards, panel consolidation, and timekeeper controls all operate after a matter has left the building. They are enforcement mechanisms, they work, and they have a natural ceiling. Once block billing is caught, unapproved timekeepers are rejected, and rates are negotiated to market, the recoverable percentage stops growing. Most departments reach that ceiling within a few billing cycles and then hold flat.

The second lever behaves differently. The Adecco Group reports a 5% to 10% reduction in reliance on outside counsel after deploying Harvey across a Legal and Compliance department of 370 professionals spanning 45 countries, supporting more than 300,000 contracts a year. That reduction did not come from tighter invoice review. It came from work that never went out.

Andreas Vosskamp, who leads Group Corporate Legal and Global Legal Operations at The Adecco Group, describes the mechanism plainly. The company does not have lawyers in every country, so specialty questions went external by default. Harvey now gives the team a first view on those questions, which means legal research, first-draft analysis, and jurisdictional comparisons that used to be outsourced stay in. AI for legal drafting does the most visible part of that, since a first draft is what a firm would otherwise have billed for.

Read that against an ELM dashboard. Every one of those referrals would have been recorded perfectly. Matter opened, firm assigned, budget set, invoice received, guidelines applied, spend reported by practice area and business unit. The platform would have described the outsourcing decision in complete detail and done nothing to change it, because the decision was made before the matter existed as a record.

Talanx Group shows the same pattern on a different workflow. The European insurance group, with 28,000 employees serving clients in 175 countries, saved more than 400 external consultant hours in 2025 by moving review of its information and communications technology contracts in-house with Harvey. Four hundred hours of external work is a line item that simply stopped appearing.

This is why the RSGI finding mentioned previously matters. When over a third of in-house teams say they are insourcing more or instructing fewer firms, they are describing a change in the volume flowing out. Volume is the variable your billing guidelines cannot touch. The enforcement lever operates on price. The insourcing lever operates on quantity. Only one of them has room left.

What Insourcing Does to Your Matter Taxonomy

When the shape of outside counsel work changes, the reporting built on top of it goes stale, and most ELM taxonomies are built around a send-it-out model.

According to the report, The Accelerating Impact of Legal AI, in-house legal departments are changing shape without changing size. Headcount stays flat for 68% of in-house teams, while 48% expect legal to take a more influential strategic role in the business. The hiring points the same way, with 36% adding legal engineers or legal operations professionals and another 36% standing up a center of excellence or an alternative delivery capability. Both are investments in internal delivery capacity, which is where a department puts money when it expects to keep more work. The in-house sample here is 27 teams, so treat the precision loosely, though the pattern matches what the real-world examples show.

A matter taxonomy designed three years ago will not describe that department well. Categories tend to be organized by practice area and by firm, because the original reporting question was which firm handled what and at what cost. When a growing share of work is completed internally with AI assistance, several things stop resolving cleanly. Internal work carries no invoice, so it generates no spend record and often no time record either. Matters once opened specifically to route work externally may never open at all. Cycle time on internal work has no billing entries to measure against.

Three adjustments are worth making before the drift compounds:

• Add a delivery-method attribute at the matter level, so every matter records whether the work was completed internally, externally, or split between the two.
• Track first-pass origin separately from final ownership, since the Adecco pattern involves internal first drafts feeding a narrower external engagement.
• Baseline referral rate by matter type now, while the send-out pattern is still the norm, because that baseline is the only way to evidence the change later.

None of this is an argument against ELM. Your organization still needs matter inventory, spend enforcement, and a defensible reporting line to finance. The point is narrower. A platform reporting on external spend will describe a shrinking share of the department's actual work, and nobody tends to notice until a board question exposes the gap.

Verification Decides What Stays In-House

The reason work leaves an in-house department has never been purely capacity. It is confidence. A lawyer sends a jurisdictional question to a local firm partly because there is no time to research it, and partly because a firm's answer carries accountability that an internal guess does not.

Insourcing therefore turns on verification more than speed, which is the bar AI for General Counsel has to clear. Output a lawyer cannot check produces no reduction in outside counsel spend, because the lawyer sends the question out anyway, or sends it out afterward to confirm.

Harvey is built around that constraint. Answers return with the authorities and source passages attached, so an in-house lawyer can open the underlying provision and check it against their own standard before anything leaves their desk. Harvey’s knowledge sources cover legal, regulatory, and tax questions across jurisdictions, which is the exact category of work Vosskamp of The Adecco Group described sending out for lack of local coverage. A qualified lawyer must review AI-generated work before anyone relies on it, and that review is the mechanism itself.

Grounding matters as much as citation. Generic AI answers a question in general terms. Harvey works against your organization's own precedent, positions, and prior matters through Vault. That puts legal knowledge management to work inside the draft itself. A first draft reflects the positions your team already takes, and a contract review flags what sits outside the standards your organization applies. A lawyer verifying generic output has to check both the law and whether the answer matches house practice. Verifying grounded output collapses that to one question.

The proof shows up in what departments report. Bridgewater Associates has described 95% time savings on large-scale agreement reviews, with one contract review dropping from an average of two days to two hours. Savings on that scale change what a team is willing to keep, and what a team keeps is what stops appearing on an invoice.

The Number Your Spend Reporting Will Never Show

Avoided referral is among the most valuable metrics in an in-house legal department and the hardest to evidence, because it is defined by the absence of a record.

Consider two departments with identical ELM reports. Both show outside counsel spend flat year over year. In the first, the business grew 4% and the legal team held the line through invoice enforcement. In the second, the business grew 22%, the department absorbed the additional volume internally, and spend stayed flat because roughly a fifth of the work that would have gone out never did. The second department created far more value, and the platform reports both outcomes identically.

Three measures make the difference visible. Track outside counsel spend as a ratio to a business volume metric such as revenue, headcount, or contract count, so flat spend against a growing denominator reads as the win it is. Record referral rate by matter type, meaning the share of matters of a given kind that involve external counsel at all. And log first-draft origin on the work that does go out. A matter where your team supplied the analysis and the firm reviewed it costs materially less than one starting from scratch at firm rates.

All three require a field your ELM platform probably does not carry by default. Adding them is a configuration decision, and it is worth making before the change you want to evidence has already happened.

Why In-House Legal Teams Use Harvey

Harvey works in the layer that decides what stays in. Four capabilities do most of that work for in-house teams, and each maps to a specific reason work used to leave.

Answers arrive with authorities a lawyer can open and check

Verification is the precondition for everything else. Harvey returns source passages and citations alongside the answer, so an in-house lawyer can confirm the reasoning before relying on it or forwarding it to a business partner. That check is what makes keeping the question internally defensible in the first place.

Work is grounded in your organization's own precedent

Harvey drafts and reviews against your positions, templates, and prior matters. A contract review flags departures from the standards your team actually enforces, which is the difference between output a lawyer edits and output a lawyer has to rebuild.

Coverage extends where your organization has no local lawyer

Jurisdictional gaps drive a great deal of external referral, as the Adecco team found across 45 countries. Harvey handles cross-border research and multilingual contract work, which lets a team take a first view on questions it previously had to send out by default.

It runs inside the tools your team already uses

Legal work is multi-stakeholder and lives in documents and email. Harvey works inside Microsoft Word, Outlook, and SharePoint and connects to document management platforms, so drafting and review happen where the work already happens. Shared Spaces extends the same environment to outside counsel when a matter warrants external help.

Where the Next Legal Operations Decision Sits

Most departments start looking at ELM because the work is piling up faster than they can handle it. Measuring that pile more accurately does not make it smaller. Demand keeps growing, the team stays the same size, and the budget no longer stretches to sending the extra work to law firms. The only lever left is how much your own team can get through.

That second decision is where legal AI belongs. Harvey works on the substance of a matter, so the research, the contract review, and the first drafts stay inside while your team keeps the judgment that has to stay with a lawyer. Your ELM reporting already tells you precisely what left the building last year. Request a demo to see how much of it would not need to this year.

Frequently Asked Questions About Enterprise Legal Management

What is enterprise legal management software?

Enterprise legal management software consolidates the operations of a corporate legal department into one platform. Core functions include matter tracking, legal spend management and e-billing, outside counsel and provider management, intake and routing, contract functionality, and reporting. Scope varies across legal tech products in this category, so the component list carries more weight than the category name.

Does enterprise legal management reduce outside counsel spend?

It reduces the price paid for work that gets sent out, through automated invoice review, rate control, and panel consolidation. It has no effect on how much work gets sent, which is the larger variable. Departments seeing the biggest reductions are changing both.

What is the difference between ELM and legal AI?

ELM coordinates legal work by recording, routing, pricing, and reporting on it. Legal AI performs the work itself, handling legal drafting, review, research, and analysis inside a matter. The two answer different questions, and a mature department runs both as separate decisions.

Does a small legal department need enterprise legal management?

Usually not yet, and it is rarely the first in-house legal software a small team should buy. A department with a handful of lawyers and modest outside counsel spend gets most of the benefit from published billing guidelines, a shared intake inbox, and a maintained matter spreadsheet. The purchase makes sense once recovered invoice errors cover the cost.

How do you measure work that was never sent to outside counsel?

Track outside counsel spend as a ratio to a business volume metric, record referral rate by matter type, and log whether your team supplied the first draft on matters that do go out. All three usually require adding fields to your ELM configuration before the change occurs.