Legal tech focus: Is Kira the real deal?

AI contract analysis system Kira has been on a trying journey. Having landed $50m in funding, Hamish McNicol and Thomas Alan assess how much longer that journey could last

Noah Waisberg recently threw a diamond into an audience of more than 1,000 people. It was the annual Legal Geek legal tech conference in London. Public Enemy’s Don’t Believe the Hype blared as he took the stage.

‘Hype is a scourge on the legal tech community,’ he told the crowd of legal tech enthusiasts. ‘It sets unrealistic expectations and it makes people dismiss what could be good applications.’

The real diamond was thrown among handfuls of fakes. Whoever found the genuine one could keep it, Waisberg said: you just had to be prepared to look. A nice analogy for a legal tech market full of noise and one helped by the fact the company he co-founded, Kira Systems, had just sealed $50m in private equity backing.

‘I should add that it was only a $400 Canadian diamond. But it was a nice-looking diamond – they’re very sparkly.’

The much-cited machine-learning contract analysis system can currently make a decent case for being that law tech diamond, following an expansive couple of years.

This was highlighted by the investment from New York-based Insight Venture Partners in September, the largest singlebacking for a company of this kind to date. Insight’s managing director Peter Sobiloff also joined Kira’s board as part of the deal.

It was Kira’s first external investment since its inception in 2011, and with Waisberg and his co-founder Alexander Hudek, Kira’s chief technology officer, retaining a majority stake, the Toronto-based company now boasts a valuation believed to be around $200m.

‘When you set people up to expect magic, they expect magic and so if you only deliver a 60% time saving, they’re disappointed.’
Noah Waisberg, Kira

Kira claims it saw year-on-year revenue growth of more than 100% for 2017, without disclosing figures. Employees, meanwhile, have more than doubled to exceed 100. Marquee clients including Clifford Chance (CC), Freshfields Bruckhaus Deringer and DLA Piper have come on board. Kira now has about 100 mostly law firm subscribers, including a majority of the Global 100’s top 30.

The company has made large strides under its own steam, but Waisberg says Insight’s funding allows it to make ‘more aggressive investments’ to create the product it ultimately wants.

That product will be one ubiquitous with contract review. Ambitious, and an objective that Waisberg concedes requires a lot more development of Kira’s software.

Most importantly for Waisberg, however, will be to not over-hype what its products can do right now. When you tell people your software is magic, it will get people trying it, but Kira wants subscribers.

‘Lots of firms tell us it’s very helpful, but people are still reviewing contracts. If someone’s saving 60% of their time using our software, that’s pretty amazing in one respect, but in another they’re still spending lots of their time reviewing the contract. When you set people up to expect magic, they expect magic and so if you only deliver a 60% time saving, they’re disappointed. It destroys credibility.’

Three surprises

Waisberg’s background is typical of many legal tech founders. He was an M&A lawyer at Weil, Gotshal & Manges in New York, supervising staff reviewing contracts for four years.

He soon realised that was not what he wanted to do. He also realised that most junior lawyers were spending endless hours on contract review work they were not that good at and hated: mistakes were being made and a significant portion of the cost of an M&A deal was soaked up by diligence.

He quit in the summer of 2010, knowing the problem but lacking the technical expertise to solve it, and ultimately met Hudek through a friend of a friend of a friend. Work started with the belief that any product would take about four months to get up and running. Six months later it was not and had no prospects of working well any time soon. It would be two and a half years before it got remotely close to what they wanted.

By the summer of 2013, however, people had tried the software on real deals and were asking to use it again. The sheer grind of building the software is a key reason Kira remained bootstrapped for so long.

‘Rather than being able to say that in six months it will work, we had to say that sometime in the next ten years we should get there. Not a solid pitch.’
Noah Waisberg, Kira

Waisberg recalls: ‘Rather than being able to say that in six months it will work, we had to say that sometime in the next ten years we should get there. “Oh and we haven’t got anyone to pay us for it yet either, but we think some day in the future, maybe, they will.” We didn’t think that was a solid pitch.’

Yet law firms were slowly coming on board and by the end of 2013 the business was generating revenue. The team expanded from four to eight over the next year and, significantly, Deloitte signed up to use the software: more than 5,000 people at that company now use Kira products, which as far as Waisberg can tell is the largest implementation of artificial intelligence (AI) software by a professional services firm.

The first wave of big law firms came in 2016, however, led by UK firms CC, DLA, Freshfields, Herbert Smith Freehills, Ashurst and Addleshaw Goddard. Allen & Overy followed soon after and Kira has taken residence in the Magic Circle firm’s technology accelerator, Fuse.

Waisberg says Kira had been talking to many firms for a while – long sales cycles plague legal tech – but by that stage the company had built up enough examples of successful deployment.

The business was still lean, however, with just 35 staff at the beginning of 2017, and was not well equipped to serve large law firms. As Kira bulked up to the more than 125 staff it has today, however, more clients have signed up, notably in the US. It now has users in more than 40 countries, while about 70% of staff are based in Toronto.

Kira’s clients include five of the top ten Vault 100 firms in the US – a ranking of the US’s most prestigious firms – with Latham & Watkins, the world’s second-largest firm by revenue being a particular standout. Audit and advisory firms use it too, as well as legal service providers – New Law darlings Axiom and Elevate are both Kira clients.

The software is taught to find provisions in contracts (which can be taught in different languages): Waisberg taught it the first 28 data points pre-2014, looking through contracts and highlighting, say, assignments clauses to teach the system what they looked like. By 2016, Kira had been taught 100 data points; now there are nearly 600. Customers are able to teach the software to find points as well – some have taught it 10,000 – while Kira has full-time staff constantly refining it.

Kira sells subscriptions to documents in advance, similar to a mobile phone plan. For example, a firm might buy the ability to review 100,000 contracts a year. By committing firms in advance – rather than charging on a per-project basis – they are incentivised to use it. The unit price per document comes down the more you buy. Kira chose this pricing model after early sales resulted in firms signing up but then not using and therefore not paying for it.

Kira also charges firms for its pilot process, which can last months. Kira gives firms a project manager for the pilot and again the charging incentivises firms to take the pilot seriously. It also means firms will only come to Kira when they are serious about wanting to try the technology.

‘There were three surprises with Kira,’ notes Waisberg: ‘It was way harder than we thought to build the software; way harder than we thought to get firms to buy the software; and that pretty much every business has massive numbers of contracts and they don’t know what they say, but they’d like to.’

‘Becoming ubiquitous’

Waisberg emphases the need for Kira to keep evolving and improving its product. This is more about the opportunity he sees, however, rather than concern about an increasingly competitive landscape: Luminance, for example, has grown substantially in little more than two years and rapidly attained a high profile in the profession.

There are two key aspects to Kira’s agenda, according to Waisberg. One is simply growth in law firms using the software to a point where it becomes synonymous with contract review.

‘It’s not like every firm uses this every time they review a contract; it’s spottier than that. When I think about redlining and how ubiquitous that has become with contract review, I don’t think contract analysis software is there yet. Some firms use it, but it’s not ubiquitous.’

‘Firms run tests between us and the competitors when they’re choosing, and big firms overwhelmingly choose us.’ Noah Waisberg, Kira

The other element is Kira’s sales pitch for large companies and large sets of contracts. Big firms have hundreds of thousands of contracts and they might have a database of who those contracts are with, but it is unlikely they know what they say.

Manually assessing those contracts at 30 to 90 minutes a contract is prohibitively costly, but the argument is that Kira could make it commercially viable.

‘We think as we develop the software it will get to be something where it enables people to know what their contracts say and if they could know what their contracts said, they would start to pull out information. That is a really exciting opportunity.’

He adds: ‘But to really make those projects happen we need to continue to push our software further than it is right now. It needs to be almost hands-off, finding all the details in the contracts but with a lot lighter touch than they have to have right now.’

This is where the recent funding comes in: Waisberg says he received lots of interest over the years and had been talking to Insight for a long time. Kira had been profitable, but only because it had not been good enough at spending its money on expansion.

Waisberg downplays talk of an initial public offering at this stage, instead focusing on that ‘becoming ubiquitous’ goal and how much work it will take.

‘Will every company in the world be our client in three years? Probably not. But we can build software that might make it possible for them to become our clients quickly. The opportunity is so much greater for us if we continue to improve the software.’

Kira draws praise from leading law firm tech leaders for its user-friendly nature. One top-ten firm’s chief information officer believes Kira completely dominates market share.

Waisberg says the main product differentiation is that people use the product. Its software is being used on about 2,000 active projects per month. Luminance said in November 2017 it had been used on more than 200 transactions, but it no longer uses this metric because it is not relevant to the new product lines. Its customer numbers have quadrupled since then, though.

‘I hope its high use is because our software is better, but it’s not like I get demos of the competitors’ software. Firms tend to run competitive tests between us and the competitors when they’re choosing, and big firms overwhelmingly choose us.’

He adds: ‘The main competitor is just people not using the software. It’s not like firms are an existing Luminance or whatever customer; it’s that they don’t use anything. It’s almost all greenfield.’

The next 12 months will be telling: $50m is a lot of capital to scale up. But there is a fast-growing market of companies touting AI solutions to compete with. Waisberg prides himself on what he says has been an understated, longer-term approach: the aforementioned not-promising-magic point.

Competition is one thing, but increasing market and noise may present a sterner challenge: nobody ever did find that diamond at the Legal Geek conference (Waisberg spotted it before cleaners swept it away and later gave it to somebody who answered a question he posted to Twitter).

‘There is a lot of attention paid to artificial intelligence and businesses capitalise on that. But there is a lot of sloppiness, mostly vendors who overstate stuff, and people jump to conclusions. We’ve always tried to be un-hypey: it hurts you in the short term, but helps in the long.’

hamish.mcnicol@legalease.co.uk
thomas.alan@legalease.co.uk