The Giants Who Offer “Good Enough”
The last two weeks have marked a progression in the legal tech landscape.
Anthropic entered the sector properly — not just a few lines of text in a skills file — but by releasing a Word add-in that connects to the broader Microsoft Office stack, moving data across workflows that lawyers already use. Then, just yesterday, Microsoft launched its own dedicated agents for legal teams.
This was always coming.
Now that we are here, what does it mean for all those who are touched by legal tech?
We Have Seen This Movie
There is a well-worn playbook for how generalist giants enter markets dominated by specialists: they wait.
They let niche players do the R&D, prove the use cases, absorb the rejections, and build customer education. Then, once the market is proven, they roll out something distinctly mediocre dressed up as adequate, and distribute it through everything they already own.
We saw this with Slack and Teams. Teams arrived late, slow, and feature-incomplete. Everyone hated it. Everyone used it anyway… Because it was already installed. Because procurement takes 18 months. Because it was the lowest effort for adoption. Microsoft knew exactly what it was doing. Teams was never meant to be the best communication tool. It was meant to be “good enough”. That is not a failure of ambition. It is the whole strategy.
We foresee Microsoft's legal tools will follow the same pattern. They will be mediocre. Lawyers will complain. Firms will adopt them anyway, because the path of least resistance runs straight through Microsoft 365. Anyone expecting Microsoft to build something excellent will be sorely disappointed.
(Anthropic is a different story. They want to own intelligence. We might deep dive into that another time)
Mediocrity is the Strategy
Most commentary on Microsoft entering legal tech treats "good enough" as a simple threshold — either Microsoft clears it or it doesn't, but it is not so simple.
"Good enough" is like a price tag. A price tag is a single number that collapses a complex set of variables — production cost, demand, competitive positioning — into one point. "Good enough" works the same way. It collapses task complexity, data quality, risks, and verifiability requirements into a single question: acceptable or not.
This means the bar sits at different thresholds depending on the task and depending on the user.
For summarizing a document, drafting a first pass, or generating a checklist — Microsoft will clear it.
For comparing two contracts at the word and character level, extracting structured data reliably across thousands of documents, or producing output that can be traced to an exact source — the bar sits much higher, and we expect Microsoft will disappoint users.
More importantly, we expect Microsoft will not prioritize the work required to close that last mile. Mediocre and widely distributed is the point.
This is not a criticism of Microsoft. It is admiration they can do so little and still achieve so much.
Four Bets on Survival
In the face of these behemoths, specialized legal tech companies face a choice. Each path is a different bet on where "good enough" lands for different lawyers in this highly fragmented market:
Fight head-on. Match Microsoft head on, and bet your product can overcome their distribution advantage.
Sidestep. Build features Microsoft will not or cannot do.
Go vertical. Own a specific workflow so completely that switching costs become prohibitive.
Integrate. Accept that Microsoft will own the interface, and build the tools that make it work better.
Many legal tech players have already signaled they will be integrating. The emergence of MCP servers and connectors — which plug directly into Claude and similar models — is how many legal tech companies are filling the gaps left by the giants. They are becoming part of the plumbing that survives by being connected to Microsoft and Anthropic.
Survival Is Not the Same as Winning
All four strategies can allow a technology company to survive, but only some end up winning.
Survival means finding a gap Microsoft does not fill and living in it. There will be many legal tech companies that do exactly this — retreating into niches too specific or too complex for Microsoft to bother with. That is not a bad outcome, but winning is different. Winning means owning tasks where the bar is structurally hardest to clear, and becoming the layer that others — including Microsoft — depend on.
Winning means being better than “good enough” — actually adopted, used, and perhaps even enjoyed.
We heard an adoption story recently that illustrates the difference between winning and surviving. An AmLaw 100 firm had been building workflows on a Gen AI platform for months. Hundreds of them. Lawyers across the firm deployed use cases with enthusiasm! The initial adoption metrics of these AI workflows looked strong, but a year later, most of those workflows have been forgotten. They joined the graveyard of shelfware — deployed, and then abandoned when lawyers discovered the output needed more fixing than doing the work themselves.
Nothing Has Changed At All
The market is about to get very noisy again. Microsoft and Anthropic everywhere, offering legal AI. Specialist vendors competing for attention, each claiming to do something Microsoft and Anthropic cannot.
The buyers will have to navigate this noise, looking for which tools can deliver real benefits. Some tasks have a low bar and Microsoft will own them. Others have a bar that demands reliability, traceability, and repeatability, and the convenient Microsoft option will cost more later.
Microsoft and Anthropic entering the legal tech vertical will reshape the market. It will expose some companies as thin wrappers offering very little value. It will force others to sharpen their offering. This is not a bad outcome.
Everything in the market has changed, but nothing about the work has changed.
