60 to 80% of revenue is reinvested via complex contracts.

Only 10% managers read the contracts that control their operations.

Making contracts highly relevant operations documents, rather than purely risk boundary controls for lawyers is our starting point for transformation.

Below, are some suggestions and how emerging Gen AI technology can help.

Data before process.

In our well-received article on this important mind shift towards data over process, firms leading the charge in getting more out of partnerships are moving away from chasing processes and controls to improve quality and are instead focusing on access to data first. Why? because they know the rest will then be positioned to follow. Let’s talk about a specific example in the area of performance contracting where outcomes are the chief concern.

The question worth posing is, “How might a small firm improve the operations relevance of its contracts and by extension, its operations as a whole?”

If we can create contracts that can effectively leverage past experiences (look backward) and effectively apply these lessons looking forward, providing useful tools and answers, they would become assets, not liabilities to be worked around, or put into a drawer.?

Contracts also have other data in them that often fail to make their way into the operations controls and talking points. By making the contract more relevant, this data also becomes an effective part of operations, and appropriately so as contracts often reflect management expectations as to the performance of the enterprise.

So how would we run this experiment?

Our contract creation process leaves us wanting

We should always start with the most operationally relevant part of the contract, the Statement of Work. Say we need to create a statement of work for that new partnership for a new platform or ERP solution or SaaS or outsourcing deal. What do we do? Let’s be honest. We make a hurried phone call to someone who might have a template, or the implementation partner accelerates the process by cutting and pasting 90% of what was in the LAST deal into this one. We haggle over what offends us and what happens to come to mind as discussions unfold and the result is often deemed good enough.

The document is so large, it becomes a challenge to know how to use it as an operating tool and describe it to those who must approve it. But we put managers on the hook for knowing what is in it at very low levels of detail. We set up review sessions where we take hours of valuable operations time reading 100-page contract documents line, by line. Why?? Because the business is supposed to know their business, so reading this document and agreeing, and getting it signed is how we make them all parties take accountability.?

They should know their stuff!

But the contract is written by a limited number of people, in a format only legal could love, and not by the ones that live in the operations talked about in this document, day in and day out. Usually, a well-meaning manager who has heard the war stories is the proxy for the team as the drafting and review cycles roll on. They do their best.

What’s wrong with this drafting process?

The drafting process itself process seems like the right place to start. If comes from the mind of the experts it has to be good right? But it is also flawed. A common known fact is how imperfect contracts are from the start, largely because projects themselves are complex and the full scope is only known at the end.

Therefore, any plan to govern complex projects is also imperfect. So most efforts are directionally accurate at best, largely aligning rules of thumb and estimated skill sets toward where it appears they will be needed based on past performance.

Most large deals emerge from the experience of consultants or implementation partners who have a strong technical perspective but are also clouded by the lack of clarity on the complete picture of how our organization works. They are also biased as to how deals should operate and be structured, opting for bare minimum resources to keep costs low We don’t design for success, but rather to avoid outright failure.? They are comfortable waiting for the "uh oh", moment later, and then simply asking for a change order.

Add to this complexity the fact that the drafting process suffers from some systemic risks by its nature.

1. Time is limited.? The teams usually don’t have time to just talk and share ideas.? Largely because the drafting process itself is so laboriously long and tedious. Also, the process to fold in outcomes from such sessions would be usually difficult and add more time pressure. So best guess efforts fill the void as they are faster and directionally accurate.
2. Vendors have the pen. We are usually recycling what someone else did before as the starting point OR having the vendor draft the technical details for us because we lack a framework to lay out the operational requirements, usually from a lack of time but also, perhaps from a lack of certainty of how to organize the work to our best advantage operationally and from a systems perspective of how our tools and platforms tie together etc. Vendors have the option to game SOW scope and structure to their benefit and we are anchored into their thinking from the beginning.? Often the program managers struggle with how to approach managing this work effectively. Often the contracts are not clear on this and offer no management roadmap beyond standard SLAs.? ?
3. No repeatable framework. If your contract writing capability were a firm, would anyone pay to use it?? For all of the years of work under our belts and expertise from all the contracts written, capabilities and tools are not evolving. We can’t say we are any better at it from one year to the next. We don’t have any IP, frameworks or artifacts, other than some standard clauses that show we have done this before. We are not capturing larger perspectives for how we guide and facilitate the business objectives into a commercial agreement structure.? We can’t repeat the wins easily outside of the personal experience silo and we can’t challenge retained team members to improve beyond personal efforts.

HOW AI Changes this Game

Generative AI allows us to easily and systemically fuse risk management into our contract drafting process, providing us benefits beyond mere speed, but also for quality of operational outcomes.?

I propose that our experiment explore how we might train AI models on outcomes of every project that we do in the firm. By creating a clear and accurate data set on contracts and their impact or lack of impact on operational challenges, we can create a data set of drafting cues that could be easily followed and extended across different teams.

Some opportunities that might be within striking distance of a trained AI are...

• Drafting with a knowledge of which processes are weak in your organization
• Mitigating what skill sets are lacking (soft and hard)
• Recall what type of reporting worked and what didn't and how that information should be used
• Flag risks buried in the work breakdown structure with counter-measures proposed for discussion
• Identify and resolve missing steps in key processes, architecture documents or deployment sequences
• Use a catalog of mistakes and past failures and create team rituals and data exchanges to minimize or eliminate them
• Leverage work breakdowns, project plans, architecture documents and process flows to outline operations inconsistencies, anomalies and risks

?Even tools and usable operational visuals can be created from this same system to make these documents more operationally relevant and less of a legal artifact. Wouldn't be nice to move from contract compliance to contract (operations) enablement.

Let me know if your firm has realized this operations opportunity, not just to speed up contracts but to improve operations by thinking about the performance data, roles, quality standards and failure types that exist within your organization. Are you using AI to pull these data points together for the first time to improve vendor operations?

Do you think this might be a worthwhile experiment?