Crosby - Building an AI-First Law Firm
- Alex Baker
- 1 day ago
- 8 min read
Most legal tech startups set out to build software. Crosby chose to build a law firm.
At first glance, that sounds counterintuitive. Why build a services company when SaaS multiples are so much higher? Ryan Daniels and John Sarihan argue that the law firm structure provides the necessary raw ingredients for true transformation. As with any new technology but particularly in the risk averse legal domain, the jury is still out.
Their approach raises questions: can such a hybrid model scale profitably, and is it replicable beyond this founding team? For now, Crosby is a fascinating, high‑cost experiment testing whether starting from scratch is the only way to reimagine how legal services are delivered.
Should others follow? Possibly, but the balance between innovation, regulation, and commercial viability is plausible but unproven.
Why a Law Firm, Not Just Software?
The traditional law firm model has always been about human capital - hiring the best talent, training them, and billing for their time. But that model has limitations:
Partnerships can’t easily invest in technology that typically delivers a return outside of the financial year and profit distribution.
The billable hour misaligns incentives, rewarding time spent instead of speed.
Innovation happens at the edges, not in the core delivery of the service and operating model of the firm.
Crosby is taking a different approach that combines engineers and lawyers in one environment, literally sitting side by side and creating feedback loops that traditional legal tech can’t capture. Every contract negotiation becomes both service delivery and product refinement.
Contracts as the “API for Business”
Crosby focuses on contracts: NDAs, MSAs, DPAs, and other B2B agreements for fast growing tech companies. These are the arteries of their clients businesses, yet the process of negotiating them hasn’t fundamentally changed since the word processor.
The Crosby value proposition is simple:
Faster turnarounds - median review times under an hour, with a roadmap to minutes.
Deal velocity - predicting which terms matter and cutting out needless back-and-forth.
Aligned incentives - no billable hours, only per-document pricing and cheaper than you will get anywhere else.
The goal isn’t just efficiency. It’s helping clients grow faster by accelerating the deals that underpin their businesses at a more economically viable price point.
Lawyers and AI, Together
Crosby has a few design choices that whilst not unique, is one of the early adopters of the AI native law firm model and in that regard, makes them worthy of exploring further.
Lawyers stay in the loop - providing oversight, liability cover, and the “taste” that machines can’t replicate (yet).
AI agents mirror law firm roles - from a paralegal-level routing agent to planned “senior associate” agents handling complex review.
Per-customer tuning - rather than chasing general-purpose accuracy, Crosby aligns models with each client’s preferences and risk profile.
By combining AI scaffolding and legal expertise turns qualitative judgment into quantitative metrics. For example, Crosby tracks "TTA" (total turnaround time) and even measures “hurt” (human review time) to ensure automation reduces both client friction and lawyer pain.
As more contracts are processed, the service will, in theory, improve for every new client creating a client value flywheel.
The Limitations and Risks
While the Crosby model is bold, it’s also a costly endeavour. The firm has raised a second round of $20 million funding (bringing total funding to ~$26m), which suggests investor confidence but also implies significant burn. This might sound negative , but it is all relative to the potential opportunity they are pursuing.
Building an AI-first law firm from scratch means carrying the costs of both a law firm and a tech company:
High-calibre lawyers alongside expensive AI engineers.
Compliance, insurance, and regulatory costs of operating as a law firm.
Continuous compute and model refinement expenses.
A long feedback loop before clear product-market fit or scalable profitability emerges.
Even with AI efficiency, the service-heavy model limits gross margins compared to pure SaaS. Unless Crosby can achieve meaningful automation leverage - moving from lawyer-led, AI-assisted to AI-led, lawyer-supervised - profitability will remain elusive.
By being a law firm, Crosby gains trust and credibility but loses the ability to scale like a product company.
Revenue Analysis: Volume to Achieve Break Even
At these volumes, and assuming additional resources weren't needed to service the demand, Crosby would need around 120 clients processing roughly 240 SaaS agreements per year at an average of $200 per document to reach an estimated break-even point of $5.76 million in annual revenue.
This provides a useful benchmark for assessing sustainability relative to the firm’s operating costs.
Volume to achieve break even | |
Clients | 120 |
SaaS agreements per annum | 240 |
Price point per agreement | $200 |
Revenue | $5,760,000 |
At this stage the model is of course unproven much like any new venture but they do have a delicate balance to strike between being too operationally heavy for venture-style scaling and too tech-forward for traditional legal economics.
That said, with investors like Sequoia and their portfolio of investments who match Crosby’s ideal customer profile, they may have the distribution on their doorstep and a clear path to their first 100 clients.
The Market Opportunity
If Crosby can prove that an AI-first law firm can scale profitably, the upside is still significant - even within its (current) narrow focus. Crosby isn’t targeting the entire legal services market. Its sweet spot is contract negotiation for fast-growing technology companies, a segment characterised by high deal volume, tight turnaround expectations, and receptiveness to new technology. In this market, time equals growth. A delayed contract can slow sales and revenue recognition.
If Crosby can meaningfully compress contract turnaround - say, from two weeks to one day, that acceleration compounds across hundreds of agreements per client per year. Add to that, every additional dollar in revenue for their clients has a multiplier in business and investor valuation, and the pitch is compelling.
For a SaaS company closing 20–30 enterprise deals per month, even a 10–15% improvement in turnaround speed could translate to millions in additional annual revenue, business value and faster cash conversion.
In short, Crosby’s model could unlock value in three ways:
Deal velocity as a growth multiplier - helping clients close faster and recognise revenue sooner.
Predictable pricing - replacing hourly uncertainty with fixed, scalable transaction costs that fit startup budgets.
Data network effects - every contract reviewed improves the system’s ability to benchmark terms, risk tolerance, and negotiation strategy across clients.
Based on some assumptions as outlined below, the serviceable addressable market for Crosby’s focus area - contract negotiation for high-growth technology companies is still substantial.
Crosby’s ideal customers are venture-backed SaaS and tech businesses that require rapid, repeatable contract negotiation for NDAs, MSAs, and DPAs. Statista estimates there are roughly 30,000 SaaS companies globally, but not all are at a stage where high-volume contracting justifies outsourcing negotiation.
Based on Crunchbase’s stage data (U.S. proxy) - where approximately 23% of startups reach Series A or beyond - the realistic global addressable pool for Crosby is around 7,000 companies.
If Crosby were to capture between 5–20% of this market, serving clients who each process 200–400 contracts per year at an average price point between $150–$250 per contract, the business could potentially generate the following revenues:
Parameter | Conservative | Mid | Aggressive |
% of market captured | 5% | 10% | 20% |
Addressable clients | 385 | 770 | 1540 |
Contracts per client (per year) | 200 | 300 | 400 |
Average price per contract | $150 | $200 | $250 |
Estimated Annual Revenue | $11,550,000 | $46,200,000 | $154,000,000 |
Elasticity and Scale
The other driver is price elasticity. If Crosby can offer a service that is both faster and significantly cheaper than traditional human-led law firms, it could unlock another portion of the market. Today, using law firms for legal negotiation for tech companies is priced out of reach for early stage startups. If Crosby’s per-document model delivers the same assurance at a significantly reduced price point,, it could expand the addressable market.
If Crosby’s AI-first model halves the cost of human-led contract review - not just through the competitive pricing but also through the cost of engaging with their service vs a traditional law firm service offering, then demand could expand by 2–3x as smaller companies access services once priced out of reach. This would bring the broader market of 30,000 SaaS companies into play.
If successful, Crosby wouldn’t just make legal work more efficient - it would become a key part of the commercial growth in the startup economy, an intelligent legal operations layer sitting between sales, finance, and legal.
The question would then be - which area of law could they replicate this model into next? This is not something that the team are talking publicly about but would feel like a natural next step.
The Future of Legal Work?
Crosby’s bet is that they can transform their commercial legal services into a “services as software.” In their vision, the lawyer of the future has removed the repetitive contract reviews and is only there for escalation of the most complex or high risk issues. And at some point, once the process and nuances have been defined in their system, they may not need lawyers at all.
We love the proposition and the opportunity it affords the wider industry to learn from it. Regardless of whether you take the Crosby approach or an alternative, there is no doubt that the lawyers who will thrive moving forwards will be the ones that develop AI services using their data, knowledge and client base.
This is a capital-intensive strategy, and Crosby is among the first to test this kind of model. The rewards could be substantial if they succeed, but for established firms, there are alternative paths to potentially achieve similar outcomes without committing tens of millions to an experiment of this scale.
Alternative Models for Existing Law Firms
For established firms, the question isn’t whether to adopt AI, but how deeply to integrate it into their business model. From our conversations with the market, we are beginning to see three paths emerging, each representing a different balance of risk, control, and return.
Evolve Existing Services Utilising AI
Use AI and automation to handle parts of the process - the repetitive, lower-value tasks that slow lawyers down.
This model focuses on efficiency: shortening turnaround times, improving accuracy, and freeing up lawyers to focus on higher-impact work.
It’s the simplest, lowest-risk way to introduce AI without changing the firm’s core model.
Build Focused, Productised Offerings
Develop client-facing, productised tools that deliver real, standalone value in specific practice areas.
These provide tangible utility and a faster, more predictable experience for clients while keeping the firm embedded as the trusted legal authority behind it.
Such tools attract new clients by providing immediate, accessible value, and create natural pathways into higher-value advisory and bespoke legal work.
This is the first step from service delivery toward solution delivery.
AI Ventures
Create or co-own AI-native ventures that reimagine how legal work is delivered from the ground up. This is the Crosby play but spun out from an existing firm, leveraging the firm's client base and knowledge to build the new venture.
It moves beyond improving an existing service and looks at a new business around a defined legal use case - combining legal IP, AI technology, and commercial execution.
The key difference is that the venture has its own brand, technology stack, and growth trajectory, (and often legal entity) allowing it to capture technology-style margins and valuations outside the constraints of the traditional partnership model.
It’s the boldest approach, but for firms willing to explore this new direction, it offers exposure to entirely new revenue streams and long-term enterprise value.
Get in touch
If you are exploring how to build an AI native firm, AI‑driven services, productise legal expertise, or experiment with new delivery models - get in touch via our contact page.
TechCrunch – “Sequoia-backed Crosby launches a new kind of AI-powered law firm,” TechCrunch, 17 June 2025.https://techcrunch.com/2025/06/17/sequoia-backed-crosby-launches-a-new-kind-of-ai-powered-law-firm
Artificial Lawyer – “Hybrid AI Law Firm Crosby Raises $20m – Cooley Invests,” Artificial Lawyer, 8 October 2025.https://www.artificiallawyer.com/2025/10/08/hybrid-ai-law-firm-crosby-raises-20m-cooley-invests
Comments