Quick Win Tip #5: Dig Deeper on Contract AI

Contract AI is not a “magic button“ that solves all your problems, but it can deliver significant benefits as long as you have a sound resourcing plan.

In the previous post of this six-part series, I discussed the fourth key tip to quick wins in contract management: leverage execs to set strategic focus, update weekly, and relentlessly prioritize. I showed you how to drive value with best practices in organizing conversations for executives with a simple template. And explained the importance of cross-functional partnerships and how to zero in on key objectives.  

Now, in this closing post, I’m going to share my fifth and final key tip to finding rapid value in your organization’s contract management initiative. I’ll also tell you how long it takes to get up and running and how Pramata can help. 

Dig Deeper on Contract AI 

Let’s look at contract AI and its impact in the post-signature world. AI delivers significant benefits if you have the right expectations and understanding. Even though Pramata includes powerful Contract AI, it’s not a “magic button “that solves all your problems.

When you look at contract AI, specifically in the post signature space, many people think it is just “legacy contract migration.” Scrape data and load it into a system, right? The reality is that’s not sufficient if you want ongoing, accurate insight and to drive adoption to non-contract experts truly. So that’s the lens to use when you’re evaluating or testing AI. Some contract AI tools even offer, “give us one contract and we’ll show you our system.” But that isn’t real-world volume and often doesn’t include non-standard or bad quality documents. So it’s important to look at it from a real-world lens: 

  • Will a salesperson use the output? 
  • How accurate is the output? 
  • Will it be up to date? 
  • Can it get things organized in a way our business needs? 
  • And if not, what resourcing do I need to apply to get it there?

Checking for contract accuracy

If you’re like most of our customers, your real world is a little gnarly:

  • You’ve got multiple repositories (or none!)
  • There are highly negotiated and non-standard terms. 
  • Some counterparties have dozens of amendments or orders. 
  • You might even have different templates from acquisitions. 

And so, to assess the impact, you have to figure out how it will perform with that gnarly real-world data, which comes down to accuracy.

The data in your contract repository has to be highly accurate. AI alone is only 30%-70% accurate. The truth is to get to 99%+ accuracy and have complete, reliable contract visibility, AI alone simply isn’t enough. No matter how much training the technology has, AI will never be able to clean the data in case of poor OCR quality documents or identify amended terms that need override. So it’s important you get organized around what resourcing will be required to deliver success.

For example, if you’re looking through the lens of a salesperson, trying to find the right contract or renewal date, what’s the effort to get the data? Can they trust it? And what’s the action for that end-user to understand the data? What is the real-world situation?

Testing with real-world volume

So, how should you test Contract AI? We recommend you:

  • Pick 50 or 100 counterparties. 
  • Pick around 100+ files if possible. 
  • Make sure there are amendments included. 

We see most AI tools start with NDAs. But unless that’s your top priority, I would focus on contracts and insights that will make the most significant impact in the real world. Test out amended terms. Is the data up to date? Can it calculate the renewal dates?

It’s essential to build a resourcing plan against these objectives. We’ve seen companies quickly realize, “Oh no, I need much more resourcing for this to be successful than I thought.” And suddenly, their budget is all used up. Or they accept lower quality data, which ends up biting them with low adoption. 

To wrap up our final tip, remember AI isn’t a “magic button” and to get to the accuracy you really need, it will always require human review. But taking the time to conduct real-world tests and having a sound resourcing plan, will ensure you find the right Contract AI solution for your organization. 


How Pramata can help

“But we also want to zero in on what’s the optimal time to value, right? What’s the approach that’s going to help the organization get value from the contracting initiative?” -Scott Curtis, Deloitte

I may be biased, but here at Pramata, we’ve designed a best-in-class approach to solve your contract management challenges, helping you zero in on optimal time to value, just as Scott suggests. In addition, Pramata has fifteen years of experience overcoming these issues and ensures ROI by:

  • Giving you everything with one simple subscription and intuitive user experience.
  • Delivering a powerful, all-inclusive repository.
  • Ensuring what you have is adoptable, even by teams outside of Legal.
  • Integrating about 20+ AI technologies with our Expert Assist team, cleansing and organizing every contract, and ensuring 99% data accuracy.
  • Getting your organization to value in as little as 30-60 days.

With those capabilities, you’ll be in a great position to succeed. If you’re ready to dig deeper on Pramata, book a demo now.

Subscribe to Our Legal Impact Newsletter

Get exclusive event invites, peer best practices and the latest industry news right in your inbox!

More To Explore