Knowable has been working on its NLP-driven contract analysis platform for several years. Backed by LexisNexis since mid-2019, the company now has more than 300 staff and a suite of tech ready and able to meet the corporate world’s data visibility needs.
As Alec Guettel, CFO of Knowable (pictured), explained, since its Axiom days, before the spin-out, it was clear that technology plus skilled human reviewers was going to be the answer to the need for companies to actually know what is in their contracts and what they mean, both legally and financially.
The challenge was two-fold, many clients didn’t even know where their contracts were, and when it came to the tech, Axiom wasn’t there yet with the NLP and machine learning capabilities it needed.
But, things have changed. With the support of the giant LexisNexis they’ve been able – without any other external funding – to get to where they needed to be. Now they are in effect going into battle with other contract analysis companies that are focused on the corporate sector.
Guettel noted however, that they are not trying to become the next CLM platform. The goal is to help corporates truly gain access to the data in their contracts and then create real business value from that information, and that alone.
‘We think it’s inevitable that everyone [in the corporate world] will access the data in their contracts, as it’s where their most valuable data is.
‘When we were part of Axiom we would do review projects for clients, e.g. for M&A, and sometimes the client would say: could you do this for ALL of our contracts? It was an epiphany.’
But, as noted, there was a challenge. They didn’t have the tech yet to really provide what they wanted. A genuinely useful digital dashboard that has imbibed 10,000s, or even 100,000s, of contracts across a company’s various divisions and national bases, was going to be hard to make with just manual labour. Then came the joint venture with LexisNexis, which had bought RAVEL and LexMachina, and developed plenty of NLP know-how.
Also, Nik Reed, co-founder of RAVEL, joined Knowable in late 2019 as well. Now all they needed was a huge amount of tech development work and to train up subject matter experts and their NLP to achieve what they wanted.
But, now they are there. And it’s great timing. The inhouse world, perhaps because of the deluge of NLP-driven CLM companies and pressure from C-suites to be more than a cost centre, are tuning into the idea that they can have access to key business data across their galaxy of contracts in a meaningful and practical way. Knowable, supported by LexisNexis, clearly wants a significant slice of this market as companies open up to the idea.
‘We didn’t have the infrastructure in relation to data analytics before, but Lexis really helped there. We then went into ‘market making’ mode and have spent a lot of time with customers,’ Guettel said.
He explained that the reality was that most companies’ contracts are ‘in varying states of disarray’.
‘For an inhouse counsel even a simple question: can a customer terminate early and will they therefore owe us an extra fee? Can lead to a wild goose chase.’
‘Some of our clients have told us they get up to 70,000 questions like this a year across the business globally, it’s an incredibly inefficient process,’ he noted.
Indeed, in fact it turns the inhouse legal team into a sort of library service, where they are constantly digging through the archives to find information that should be at the tips of everyone’s fingers.
What Knowable Now Does
So, what does Knowable do to help? Guettel explained that the first thing they do is put all the company’s docs into different families. This, he said, is vital. It’s the foundation of all that comes next.
Then they sort out what contracts are active vs inactive, again, a hugely important step in making sense of your contract stack.
Then they build in searchability, and then finally they get into detailed data analysis around key clauses, e.g. terminations and renewals.
This is a massive piece of work, he added. As well as having their own data scientists, they also have teams across the world from the US to the Philippines, who focus on certain parts of these contracts. With this huge effort they now feel they’ve sufficiently got on top of the problem to be really useful to companies.
‘We have been working on this for five years,’ Guettel stressed, and that just underlines how much effort is needed to make an NLP tool for complex legal documents actually useful.
And, as he also stressed: there is no point offering clients results that are not accurate. 80% is OK perhaps for a fast due diligence review, but when a CEO wants to know exactly how many contracts in the UK are affected by X or Y terms, they need a correct answer. Hence you need plenty of human input to get there. Although, he added, year by year their NLP is getting better at the same time.
But, human reviewers will remain vital. ‘We have human specialists working in 30 languages. We have a team that just does indemnity and liability clauses!‘ he added, and mentioned that the 300 staff figure is not the ceiling, more staff are being hired all the time.
What Next?
Perhaps this shows that sometimes slow and steady can put you in a very competitive position. But does the market really want this? Although contract data invisibility is a pain, companies have lived with it for decades. Why change now?
‘The market has really evolved recently,’ Guettel explained. ‘Look at it this way, 90% of your revenue and risk is in the executed agreements….not the new ones.
‘Also, your old agreements will help with new ones, as many of the deals are with the same customers. You can see what positions you took across issues,’ and much more, he added.
He also noted that CLM tools cannot easily do what they are meant to do. There is just so much data there, so many contracts and different versions of them scattered across a large multi-national company.
Many companies also don’t even have just one CLM, with some having multiple CLM systems for different parts of the company. Others have no CLM or contract management system at all. It’s a messy world and Knowable has been diving head-long into it to try and build a solution, which is not a multi-faceted CLM platform, but just focuses on one core problem: what is in my contracts?
The plan now is to roll this out across the world, providing companies with the kinds of insights they can actually leverage.
Conclusion
Artificial Lawyer then asked what is Knowable? How would you define it? Guettel replied that: ‘We are a hybrid. We are not really an ALSP, and not really just a tech company.’
But, either way he and Mark Harris, CEO of Knowable, are excited about the future.
‘GCs have been in a defensive posture for 20 years, but now we can show how they can contribute to the profitability of the company. It’s about turning the legal team from a cost centre into a profit maker,’ he concluded.
And this is indeed an exciting prospect. They have arrived with a substantial product at just the right time. Now all they need to do is compete with a growing number of CLM companies and inhouse-focused contract analysis tools that are also seeking to provide the same benefits. But, after years of focused effort on these problems, and with the backing of LexisNexis, they have every chance of succeeding.