top of page

Is LegalSifter a Minimum Viable Product?

When I joined LegalSifter as chief content officer after having been on my own for fifteen years, I entered a process-driven world. I’m now part of a production line that combines technology and expertise to create software and advice. There’s lots to keep track of, so we’d be lost without our databases.

But I haven’t taken the time to become familiar with the broader conceptual framework. Instead, I’m privy to discussions in which words like agile, lean, kaizen, waterfall, scrum, roadmap, kanban, and others are bandied about, leaving me not much the wiser. But I made an exception for the phrase minimal viable product, or MVP.

I looked into the concept of MVP after hearing people use it in a manner that doesn’t inspire confidence—as an excuse. A product is lacking in some fundamental way? Well that’s OK, it’s an MVP! They’ll work out the kinks as they go along!

Instead of deciding for ourselves what MVP means, it makes sense to start with a definition. The notion of MVP appears to arise out of the lean startup world. Here’s the definition popularized by Eric Ries:

The minimum viable product is that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort.

Regarding the notion that MVP is a license to build something basic, even flawed, here’s what Eric says:

MVP, despite the name, is not about creating minimal products. If your goal is simply to scratch a clear itch or build something for a quick flip, you really don't need the MVP. In fact, MVP is quite annoying, because it imposes extra overhead. We have to manage to learn something from our first product iteration. In a lot of cases, this requires a lot of energy invested in talking to customers or metrics and analytics.

Here's another take, from the consulting firm Slalom:

As this definition makes clear, the MVP is not a product with the least possible functionality necessary for a public launch. It has nothing to do with publicly releasing a product at all. Rather, the MVP is the key to using the scientific method for building products. It is purely a mechanism for validated learning, used to test hypotheses and discover what will meet customers’ needs.

OK! Now we’re getting somewhere. But once I was able to safely discard the MVP-as-excuse rationale, it caused me to wonder whether MVP is relevant to LegalSifter. I’ve concluded that mostly, it isn’t.

That’s because we’re not working with a clean slate. Instead, we’re faced with making sense of a mass of chaotic data—the world’s contracts. That requires understanding deals, the real and not-so-real issues underlying deals, the good and not-so-good ways people try to express those issues, and what makes sense, in terms of deal points and the language used to express them.

Mostly, that work doesn’t require validated learning from clients. Instead, it requires data scientists, contracts specialists, and processes that allow those teams to combine their talents and work effectively. The one exception is when clients roll up their sleeves and help us produce Sifters (our algorithms) and advice for a particular kind of transaction—we’re happy to have their expertise. But in that capacity, they’re not really clients; they’re more teammates. Otherwise, clients are primarily in the unenviable position of being prisoners of a dysfunctional contracts ecosystem.

What clients can mostly help us with is refining the user experience. And their needs determine where we turn our attention. But that’s not particularly challenging, compared with producing Sifters and the advice that accompanies them.

So no, MVP isn’t something I’ve heard anyone at LegalSifter talk about.

But let’s return to what people think MVP means. In that regard, sure, LegalSifter started business once those in charge thought they had something to sell. Given the task, necessarily LegalSifter’s first offerings were minimal. But that’s OK—given the monumental nature of the task, you have to start somewhere. Everything since has involved incrementally building on that starting point, and that’s the way things will continue.

And thankfully, we haven’t had occasion to trot out MVP as an excuse for screwing up.

Recent Posts

See All

Recently I noticed this article on Artificial Lawyer. The title is Generative Legal AI + “The Last Human Mile”, and it’s about limits to applying AI to legal work. It says this: The last mile problem

bottom of page