by Jim Wagner
The More Things Change …
When I was a kid, my dad used to tell me (true) stories about growing up in a house without plumbing or electricity. I’m afraid I don’t have a lot of those tales to tell my kids (maybe, “I didn’t get my first video game until I was 10”), but if they cared to listen, and they do not, I do have a few “in my day” stories to tell about when I was a young lawyer going to vaults and warehouses to conduct deal diligence. Back in the ‘90s, diligence meant “boxes of documents.” If you wanted to find a contract, you looked in the file folder. If you wanted to look for a specific topic in the contract (say, for example, the words “change of control”) you couldn’t run an electronic search for the magic phrase; you had to review the document line by line. About the best thing we had going was sticky notes. And the closest you came to “machine learning” was keeping a log of your documents and issues list in a spreadsheet.
“Wait a minute,” you say with a wry and knowing smile. “Are you describing your experience from 25 years ago? Or are you describing my visit to online data room X yesterday?” Ah, perhaps my homey tale is a little too transparent. Because, the irony is, 25 years after my first diligence project, very little has changed. Sure, today you “go” to data rooms on your computer instead of going to a basement in lower Manhattan, but structurally and philosophically most online data rooms simply replicate the end-user functionality of the paper-based diligence that has been taking place for decades. Leave aside the real possibilities for where and how the legal diligence process can and should go (ahem, see below), don’t you find it odd and even a little offensive to go to an online data room and not be able to run a keyword search in a contract? Talk about a time warp. And, what’s worse, what an outrageous waste of time and money.
It’s clearly time for a change. Or more accurately stated, changes.
Going Beyond the Basic
At the end of last week, we announced the release of M&A2, Apogee’s proprietary analytics package designed to automate a substantial portion of the mergers and acquisition contract diligence process. The basic concept behind M&A2 is pretty straightforward – take the best of machine learning tools (think IBM’s Watson), train the platforms using experienced transaction professionals, and apply the output to the analysis of corporate contracts. And while the basic concept behind M&A2 is simple, the strategies for implementation, automation, and the range of possible use cases are truly exciting.
As I referenced above and in an earlier blog, I began my career as corporate attorney, negotiating mergers, acquisitions, investments and offerings for clients ranging from family-owned businesses, to private equity firms, to global public companies. Despite the difference in size and complexity of the various deals, the contracts diligence process was largely paper based and consistently addressed a relatively narrow range of common elements and questions:
- Change of control and assignment
- Renewal and cancellation rights
- Exclusivity, rights of first refusal and “most favored” status
- Restrictive covenants (non-solicitation and non-competition)
Topics and questions that have become increasingly relevant to sophisticated parties (cyber security, FCPA compliance, carve-outs from limitations of liability, unique insurance requirements, regulator rights, etc.) weren’t part of the equation – we simply didn’t have the budget or client mandate to spend time examining contracts for such a large variety of issues.
The irony is, despite the shifting landscape and the increased importance of understanding these “new” topics, both for risk and compliance purposes, today’s standard diligence process has evolved little. Our informal survey of, and firsthand experience, with many of America’s best law firms reflects the fact that end clients often don’t have an appetite for a more detailed diligence review beyond the basic change of control questions.
Leveraging Machine Learning to Create an Unprecedented View
At Apogee, our experience is that the inherent limitations in most online deal rooms, and the time and resources associated with the legacy diligence process, drive client acceptance of a truncated diligence analysis. Simply put, in most environments it takes a lot of time and effort to select which contracts to review and even more time to manually sift through every page to find and analyze the relevant clauses. The default, of course, is to focus on “material customers” and “material vendors.” While the logic is obvious, the reality is that even insignificant contractual relationships (financially) can contain very significant risks and obligations. The challenge, then, is to determine which contracts (whether they are material or immaterial) contain “material” provisions.
With M&A2, our approach leverages machine learning to enable clients and counsel to review every contract for both change of control issues as well as for more complex risk and compliance issues, and to do so at a lower cost than the status quo. Leveraging our experience with predictive coding in multi-million document discovery cases, our extensive backgrounds in the M&A world, and our understanding of the best machine learning tools, we created analytics that automate the identification of the most critical information (without human review of the entire contract) and do so with an exceptional degree of accuracy–finding a very high percentage of the relevant information (recall), and limiting the number of false positives (precision). Using this approach, we are able to quickly (as in … give us a few hours) and cost effectively triage our clients’ data sets and give them a “snapshot” of key M&A and risk issues. By utilizing analytics and automation, we allow our clients to make informed decisions about not only the specific contracts that they want to review, but also the specific clauses that they want to review – all at a lower cost than traditional methodologies.
Confidence in Today’s “Future”
The biggest impediment to adopting analytics as an approach to executing M&A diligence (other than inertia) is confidence. It’s clear that an automated approach can deliver results quickly, can examine a much broader range of topics, and can generate tremendous savings. The only question then is does it really work? For Apogee, this is the crux of what we do. Our expertise lies in both our substantive knowledge of the M&A diligence process, but also our ability to select and get the most out of the best available machine learning platforms on the market. That’s why we have spent countless hours partnering with the best technology players in the space and building our own custom analytics packages on top of these platforms to deliver the highest quality and most comprehensive M&A analytics solution in the market. And if that statement seems a little bold, just know that when you’ve put in the time that we have, and generated the results that we do, we can say it with confidence.