AI in business software is often hyped but rarely measured. Every week, headlines tout new AI features, but many organizations are left wondering: is anyone actually using them in real workflows? At Concord, we’ve been tracking how our AI Copilot is used inside the contract editor, and the numbers tell a clear story. Not only is Copilot being used at scale, but it’s also showing how AI adoption works when it’s embedded in the flow of daily work.

The snapshot in numbers

Since launch, Copilot has processed 7,491 queries across 392 organizations and 723 unique users, spanning more than 3,174 active documents.

On average, each user engages with Copilot 10.36 times, which shows this isn’t just a feature people try once and forget. Instead, they keep coming back to it as part of their routine contract work.

Copilot also delivers a strong performance profile. 96.9% of queries succeed, and most responses arrive in under five seconds.

These numbers matter because they mark the difference between surface-level AI usage and genuine adoption. Surface usage looks like one-off experiments or pilots. Adoption looks like sustained, repeat use at scale, with efficiency gains that change how teams work. Copilot is firmly in the adoption camp.

Who’s really using Copilot

One of the most surprising findings is that most Copilot users aren’t lawyers. While some general counsel teams are active, the majority of usage comes from legal operations professionals, finance, procurement, and business operations staff. This mirrors our broader customer base, where 70% of Concord customers don’t have in-house legal departments.

That fact points to a broader shift in contract management. For years, contracting was gated behind legal teams, even for routine tasks like checking renewal dates or confirming payment terms. With AI embedded directly in the editor, those tasks can now be handled by the people closest to the business process. In practice, that means finance teams can validate fees, procurement can track renewals, and operations can check compliance—all without waiting on legal.

This democratization is why we say it’s time to “fire legal” from routine contracting. Legal is still critical for high-stakes strategy and negotiation, but it doesn’t need to be the bottleneck for everyday contract work.

When adoption happens

Looking at time-based usage patterns, Copilot queries cluster midweek (Tuesday through Thursday) and peak in the afternoons. Usage is low on Mondays and Fridays and almost nonexistent on weekends. In other words, Copilot is being used at the exact times people are heads-down in contract review and negotiation, not as a novelty tool outside of work hours.

That pattern is important because it reinforces that Copilot is part of the workflow. Adoption spikes also occur around quarter-end and other contract-heavy business cycles. These “burst” moments of enablement are followed by steady baseline usage, showing how AI adoption follows a burst-and-sustain model: introduce the tool when demand is highest, then let repeat use settle into routine work.

What people ask Copilot to do

The majority of queries are high-intent and outcome-driven. Users aren’t asking Copilot to explain abstract concepts—they’re asking it to perform specific contract tasks:

• “Summarize the renewal clause in this MSA.”

• “What are the payment terms in section 4?”

• “Does this NDA reference GDPR?”

• “List all parties to this agreement and their obligations.”

When we categorize the queries, several clusters dominate:

• Renewals and dates: identifying expiration terms, auto-renew clauses, and notice periods.

• Termination rights: checking under what conditions contracts can be ended.

• Payments and fees: validating pricing terms, fees, and discounts.

• Compliance and risk: finding references to GDPR, HIPAA, SOC 2, or confidentiality.

• Summaries and extractions: quickly condensing clauses or pulling structured data.

These are not trivial questions. They’re the same pain points that slow down contract review and create risk when overlooked. The fact that Copilot is handling them with a 96.9% success rate speaks to how well it’s being trusted in live workflows.

Why proximity matters

One of the clearest lessons from our data is that embedded AI beats standalone AI. Copilot adoption works because the tool is inside the document editor, right where contract work happens. Users don’t need to copy-paste text into an external chatbot or toggle between apps. That absence of context-switching is what makes adoption stick.

This aligns with what we’ve seen in other parts of the market: AI features that live inside existing workflows have much higher adoption rates than separate tools. In contract management, where efficiency is paramount, embedding is non-negotiable.

Action items for AI adoption

The broader lesson here isn’t just about Copilot. It’s about what separates impactful AI adoption from surface-level usage. Based on what we’ve observed, here are four actionable takeaways:

• Measure repeat usage, not signups. Adoption isn’t about how many people tried the feature once, but how many keep using it in daily work.

• Time enablement with business cycles. Rollouts and training land best during peak contracting periods, when demand for speed is highest.

• Expand access beyond specialists. Contract AI is most impactful when finance, procurement, and operations can use it directly.

• Anchor adoption in real use cases. Focus on the clauses and questions that create real friction: renewals, termination rights, payments, and compliance.

These are lessons any team can apply when thinking about their own AI rollouts, whether in contracting or elsewhere.

The bigger picture

The debate in AI right now is about hype versus reality. At Concord, the reality is clear: Copilot isn’t just being tested, it’s being used, at scale, across functions, with measurable impact. The data shows that when AI is embedded where work actually happens, adoption follows. And when adoption follows, efficiency gains compound.

The numbers prove it. Thousands of queries. Hundreds of organizations. Near-perfect success rates. Sub-five-second responses. All at a cost of pennies. More importantly, Copilot is showing that contract management doesn’t have to be a legal bottleneck anymore. Finance, procurement, and operations teams are stepping in with AI at their side, and the results are better contracts, faster cycles, and less wasted time.

For teams still trying to separate real AI adoption from inflated usage claims, these benchmarks offer a simple litmus test: look for repeat, in-workflow usage tied to core business tasks. That’s what adoption looks like. And that’s what Copilot is delivering.