Several in-house innovators are converging on a set of best practices.
In Competition based on better commercial contract terms (211), I reviewed the current norms surrounding commercial contracting and postulated that the growing transparency regarding what is market for a particular term would cause the market for contracts to evolve from its current souk-like state to something that more closely resembles a modern e-commerce marketplace. Since that post came out in December 2020, numerous companies have been employing AI tools such as TermScout. and crowd-sourced data such as Bonterms, to make their contracting practices more data-driven.
I recently spoke with a number of leading innovators in this space, including Paul Shoning, GC – Automox, Connie Brenton, VP Strategy, Technology & Operations – NetApp, Andy Banquer, VP, Corporate Solutions – Quislex, Justin Widlund, Associate GC – FreshWorks, David Huberman, GC – Code42, and a host of large company GCs/AGCs and Legal Ops heads who asked to remain anonymous, to see how they were incorporating data into their contracting practices and what impact it was having on their departments’ ability to deliver value to the business.
The picture that emerged shows the evolution to data-driven contracting is in its infancy but that at least three best practices are beginning to emerge: (1) data-driven template optimization, (2) data-driven approach to counterparty paper, and (3) data-driven playbook optimization.
By employing these practices, the innovators have seen significant efficiency gains and expect to see even larger gains down the road as the tools and data sources continue to improve.
[Disclosure: The author is a TermScout co-founder, and all of the examples cited in this article reflect data resulting from the use of the TermScout tool. Results from the use of other tools may differ from the data reported here.]
This post chronicles each of these emerging best practices and how they are driving value to the business.
1. Data-Driven Template Optimization
The holy grail in contracting efficiency is to get the other party to agree to your standard contract terms with no negotiation whatsoever, much like what happens when someone buys an app from an app store.
When this occurs, the parties typically experience negligible transaction costs, complete their deal in minutes rather than days, and have a contract that (at least from the drafter’s standpoint) is perfectly tailored to the specifics of this transaction. From a business standpoint, this means spending less on legal, getting revenue/product access faster, and having a lower likelihood of contract value leakage. See Post 228 (discussing the often untallied costs of contract complexity).
Today’s innovators are using data and AI tools to benchmark and optimize their contracts to drive dramatically higher levels of acceptance from counterparties. While contract benchmarking and optimization have been occurring for decades, these innovators have done three things in their approach that have allowed them to improve counterparty acceptance.
Do business on “market” terms. These innovators have made the deliberate policy decision that they will do business on “market” terms, even if those terms aren’t slanted heavily in their favor. This may seem like an obvious requirement to drive counterparty acceptance, but countless examples abound where companies (especially large ones) know that one or more of their terms are out-of-market and generating friction but refuse to change course because the owner of the term (which often is an organization other than legal) remains idealistic and vetoes a more pragmatic approach. Today’s innovators have consciously chosen the pragmatic path and where benchmarking shows their terms to be out-of-market, they make the necessary changes to address this fact.
Higher quality benchmarks. The innovators are leveraging AI tools and large sets of structured data to generate higher-quality benchmarks that are convincing to counterparties. The traditional approach to benchmarking involved having outside counsel perform an ad hoc analysis of a relatively small number of agreements and produce a judgment of what is market for key terms that largely reflected counsel’s individual experience. Today’s innovators compare their terms against large sets of competitor agreements (both templates and negotiated agreements) that have been broken into points of structured data that can be compared on an apples-to-apples basis. These analyses are based upon science as opposed to art. Companies typically regard their traditional benchmark analyses as attorney-client privileged and do not share them with the other side. Notwithstanding the privilege issue, the likelihood of a counterparty giving credence to a market analysis prepared by one’s own law firm advocate is slim to none. By contrast, today’s innovators are using transparent benchmarking reports generated by an independent third party.
Independent certification of terms. Today’s innovators have learned that offering terms that are objectively balanced or even counterparty friendly is not enough to get counterparties to accept them as-is. For years, drafters have commonly slanted terms significantly in their favor, causing counterparties to approach the transaction with deep levels of mistrust. The complex nature of most contracts, which commonly are tens of pages long, include multiple sets of nested documents, and often are poorly written in legalese, further compounds the perception issue by making it time-consuming and expensive for the counterparty to even understand what is being offered. See Post 228 (discussing untallied cost of contract complexity). To address this issue head-on, today’s innovators are using data to address this issue head-on, getting their terms independently and publicly certified as balanced (or even favorable).
Below is an example of the certification based on the TermScout tool:
The combination of the above three tactics then become the basis for a comprehensive communications campaign to publicize these results, including:
- Embedding link to the independent report in the contract template itself so that the other party can easily validate that the certification is real. See Freshworks’s terms of service example.
- Educating Sales and Other Business Personnel about Legal’s efforts to reduce negotiations, including the independent certification, and explaining to them how the deal can get done more quickly on the company’s standard paper.
- Pushing Back in a standardized fashion on attempts by counterparties to use their paper or negotiate terms before they actually read the independent report in detail.
The experience of the innovators interviewed for this post indicates that taking the above-described actions can drive significant increases in counterparties’ willingness to do business on the company’s standard terms without negotiation.
For example, in the quarter following the implementation of these steps, Automox, an IT Operations cloud solutions provider, 85% of all new label deals (i.e., deals with companies that did not already have a contract with Automox) were done on Automox’s standard paper without negotiation. Automox’s GC, Paul Shoning, had expected it to take a year to reach these levels and is now looking to increase his target for the year, observing that ”this is significant for us because every percentage point we add to the number of deals we do without negotiation directly impacts the number of lawyers we need to get the company’s business done.”
Freshworks, a SaaS business solutions provider, had a similar experience. By taking the steps described above, Freshworks believes it has reduced its negotiations of sales contracts by over 30%. Justin Widlund, the Freshworks AGC driving this exercise, noted that “these changes have allowed our company to close business faster, speeding our time to revenue and allowing us to develop better relationships with our customers.”
It obviously takes two to tango and the following section shows how, consistent with the data cited above, counterparties are responding with an openness to accepting paper that is reasonable without negotiation.
2. Data-driven approach to counterparty paper
In the perfect world, all transactions happen on your paper without negotiation, but in the real world, it often will be necessary to work off of the other party’s paper, especially when you are a buyer making a relatively small purchase.
Most companies face a crushing volume of these sorts of deals, which puts the legal department in a very difficult position. You know that you lack the resources to review every deal, see Post 228 (discussing level of effort to review even a “simple” contract), but you also know that a percentage of these agreements likely contain terms that will be bad for your company and even potentially catastrophic. In many ways, the lawyer faces the same dilemma as a customs agent at a busy border crossing: “How do I decide which cars to search in order to best keep bad things out of the country without bringing commerce to a grinding halt?”
Traditionally, legal departments have taken one of two approaches. The first is to adopt a policy of not reviewing contracts that have certain attributes (e.g., below a certain dollar threshold, cover things that are not embedded in our products, etc.). This approach, however, is just as leaky as if customs were to say that we aren’t going to search any vehicle driven by someone over 60.
The second is to push the responsibility for conducting the contract review to someone else—typically the businessperson sponsoring the deal or sourcing/procurement—and expect them to be able to spot anything toxic and escalate that to Legal. Although this approach is better than nothing, it is unrealistic to expect people who are not trained in the area to effectively parse highly complex agreements. Further, this review adds time and cost to getting the deal done.
Today’s innovators have attacked this problem by using AI and big data to automate an approach of “trust but verify.” See Post 236 (describing a pilot run by multiple companies and service providers to use AI to automate low-risk contract reviews); Post 269 (Lucy Bassli including this approach as part of optimized contract lifecycle management).
First, the company must make a policy decision as to what constitutes an acceptable contract for each type of transaction (low-risk, medium-risk, high-risk) that it engages in. This requires the company to:
- Create a checklist of the terms that it considers essential
- Decide what the minimum acceptable position is on each of these terms, and
- Validate that minimum acceptable position against market data.
Once a company takes this step, it can use AI to scan its incoming agreements and receive a report showing which agreements fail and on what terms. This is somewhat analogous to the customs agent using a drug-sniffing dog to identify which bags are likely to contain contraband. Below is an example of such a report using the TermScout tool:
With these reports, companies can do two things. First, they can safely ignore contracts that pass the screen and sign them as is. Second, they can focus the attention of their negotiators (be they lawyers, contract managers, or sourcing specialists) on the terms that fail and devote no energy to the rest of the agreement.
The innovators employing this approach have seen significant results. For example, Andy Banquer, VP of Corporate Solutions at Quislex, a leading legal managed services provider, supports the sourcing organization at NetApp, a Fortune 1000 company that provides Cloud-based data management solutions. Banquer estimates that implementing this approach has allowed NetApp to increase the number of contracts it accepts with no negotiation by approximately 10 percent and to reduce the number of terms that get negotiated by approximately 25%, resulting in significant decreases in transaction cycle time.
This disciplined approach to handling third-party paper has allowed NetApp to leverage its analysis tool to largely automate the process for building an audit trail of approvals, reducing the time spent on this task by roughly 80%. While less quantifiable, this approach relieves NetApp’s sourcing specialists from having to conduct ad hoc assessments of contracts. It also allows them to focus on their core area of sourcing terms and to handle any ancillary negotiations with a predefined playbook, resulting in higher levels of employee satisfaction and a closer relationship between Sourcing and Legal. Cf Post 228 (discussing how greater attention to business terms leads to higher quality contracts). Connie Brenton, NetApp’s VP of Strategy, Technology & Operations, and the sponsor of this project observed that this approach simply would not have been possible 18 months ago. However, with the recent advances in AI and Big Data, she expects to see even greater gains in the coming year.
NetApp’s experience with using a data-driven approach to counterparty paper is not unique. David Huberman, GC of Code42, a software provider focused on insider risk protection, estimates that they have witnessed similar efficiency gains from taking this sort of approach. Automox likewise has experienced comparable gains in efficiency, along with the types of enhanced employee satisfaction witnessed by NetApp. Paul Shoning, Automox’s GC, aptly noted that lawyers are trained to make everything perfect and Automox’s lawyers have felt a sense of relief in having a data and policy-driven reason to give themselves the permission to limit negotiations to a smaller set of terms.
3. Data-Driven Playbook Optimization
We live in an imperfect world and irrespective of which party’s paper is being used, certain contracts are going to require some negotiation. Sometimes this is because the drafter takes a position that is clearly out of market. Other times it is because “market” cannot be defined with precision.
For example, suppose an analysis of the relevant set of agreements (whether templates or negotiated agreements) shows that parties agree to the following positions regarding the customer’s rights to terminate or cancel a SaaS contract with the following frequency:
You know from your experience supporting these deals that a customer’s right to terminate for convenience is an issue that comes up fairly regularly. The data shows that customers succeed in getting these rights a little over one-third of the time, meaning that the company can get away with not offering such rights, which are detrimental to the company, roughly two-thirds of the time.
Under these circumstances, one could logically decide that an event that occurs roughly one-third of the time does not rise to the level of defining “market” and that it does not make sense to offer a right to terminate for convenience as part of the company’s standard template. But one-third of the time is still a lot, and a smart company will have a standard, market-based response for dealing with the objection when it arises. The logical way to do this is to cover the issue in the company’s playbook in a fashion that allows the person handling the deal to close the issue out quickly.
The innovators that have taken this data-driven approach to optimizing their playbooks have seen significant results. For example, Andy Banquer estimates that NetApp has reduced the number of turns on negotiated terms by roughly 25% since it optimized its playbooks using market data. Automox has seen similar levels of improvement and has a goal to complete 75% of negotiations in no more than two Automox touches. Once again, this approach leads to lower transaction cycle times, happier employees, and a better relationship between Legal and its clients.
What Comes Next?
Data-driven contracting processes are showing strong initial returns but remain in their infancy. As data sets continue to grow, AI tools become more powerful, and innovative law departments gain more experience, it seems logical that data-driven contracting will become more widespread and deliver even greater efficiencies.
In the longer-term, data-driven contracting has the potential to move contracting to a Moneyball exercise where decisions regarding how to draft or negotiate a particular term are based almost entirely upon empirical data as opposed to gut feel. With sufficiently robust databases and AI tools, there is no reason that these decisions cannot be converted into financial analyses, e.g., the all-in cost of taking position A as opposed to position B on a term is $X while the probability-adjusted payoff is either $X+ or $X-. This is a topic that I hope to explore in another post.