Where the LegalTech Investment Thesis Splits: Trust, Billing & AI-Native Law Firms | Part 3/3 - LegalTech Investment Thesis Series

Venture capital firms are no longer debating whether AI will impact legal services. The deeper disagreement now lies in how that transformation unfolds, who captures value from it, and whether the future of legal services will continue to resemble the traditional law firm model at all. While the previous section identified the broad areas of convergence across LegalTech investment theses, the differences between these firms become more visible when questions of pricing, trust, workflow ownership, and service delivery are examined closely.

This Part in the LegalTech Investment Thesis Series explores the major divergences across leading venture capital perspectives on LegalTech and AI-driven legal services. Some investors believe LegalTech will remain primarily a software and workflow infrastructure market that augments lawyers within existing firms. Others argue that AI-native legal service providers could fundamentally reorganize how legal work is delivered, priced, and scaled. These divergences reveal competing theories about billing models, institutional trust, customer ownership, and whether LegalTech startups should sell software to lawyers or directly compete with them through AI-native service delivery.

The Divergences

1. Build vs Acquire: GC’s roll-up model requires owning legal service delivery through acquisitions. a16z is interested in selling legal software into the market. However, Sequoia is adopting a paradoxical approach, categorizing the LegalTech market as “co-pilot” and “auto-pilot.” Portfolio companies such as Harvey (Co-pilot) and Crosby (Autopilot) are examples of this. Sequoia still does not believe in service delivery through acquisition. a16z’s thesis offers insight into the GTM strategy that such startups can adopt. a16z emphasizes network effects (building a broad customer base), which would compound in a profession like law, where trust, reputation, and patience are key. LegalTech startups must embed themselves into law firms' day-to-day workflows, thereby becoming a default choice for those firms. 

   
   

   Battery Ventures is more interested in building solutions for the LegalTech market. This is clear from investments in startups like Ontra (an AI-powered contracts platform for funds and private-market clients). The thesis is more focused on augmenting lawyers than on acquiring or delivering AI-native legal services. Lawyers would be performing higher-value work such as case strategy and argumentation. Lightspeed is not excited about delivering AI-native services due to the operational complexity of scaling a law firm with clients and the challenges posed by the fixed-fee model.

   
   

   The TLTF fund has invested across multiple categories of software for lawyers, rather than in legal services. TLTF has a portfolio of companies across contract review, AI governance tools, immigration case management, estate planning, marketing and lead generation, citation verification, and legal research tools. TLTF reflects value in LegalTech accruing across workflow infrastructure rather than service delivery.

   
   

   This distinction reflects a fundamental difference in the bet. GC believes that owning the data moat can be achieved through controlling delivery. Other venture capital investors, aside from Sequoia, believe that distribution through existing buyers is sufficient. Furthermore, if GC’s bet on service delivery through acquisition is correct, law firms' margins would decline, and vertically integrated AI-native firms would emerge. a16z’s thesis would mean that incumbents retain client trust and LegalTech startups capture recurring ARR. Law firms would not radically change their business models.

   
   

   The main question that remains is whether LegalTech is a software market or a service. Sequoia views it as both, while GC and the other firms view it distinctly.

2. How Fast Does the Billing Model Change?: a16z does not believe in the business model changing. LegalTech startups will succeed only if partner incentives align and help increase partner profits. The argument that the billable model is changing because BigLaw firms are being forced to adopt AI is explicitly rebutted by a16z’s thesis, which holds that clients are still willing to pay high fees for high-stakes matters. Lightspeed identifies the billable model as a constraint because law firms would not be able to experience any cost-saving benefits. However, the question of whether the industry will change remains open-ended. Furthermore, it highlights that the shift from the billable to an outcome-based business model will be the key factor in determining who wins across LegalTech, Lawyers, Law Firms, and Clients.

   
   

   Battery Ventures' thesis does not explicitly mention a change in the billing model. However, it distinctly recognizes the billing model as an obstacle to AI adoption. This thesis suffers from a limitation of analyzing the billable model, which will create a hindrance for lawyers. Sequoia’s software thesis services do not include an angle on the billable model and instead focus on using outsourcing as a wedge to compound towards intelligent heavy work.

   
   

   GC does not mention the billable model, but Eudia offers subscription-based services, contract analysis, M&A due diligence, and compliance. GC’s thesis is tied to the billable model transformation if legal services become operationalized and priced under an outcome-based model rather than bespoke hourly advisory work. The TLTF believes in the Law Firm 2.0, which would include a scalable tech-services delivery model (outcome-based) and non-lawyer ownership. Thus, TLTF is adopting progressive regulatory changes, including permitting new structures under the Alternative Business Structures category. If the billing model shifts to an outcome-based one, the cost of legal services will decrease as incumbents are forced by new entrants and old entrants who adapt to the new model. If the billing model does not change, LegalTech startups will not be able to sell time-saving products because the incentives do not align.

   
   

   The core divergence is not whether AI improves legal work but whether it changes how legal work is priced. The deeper argument here is that the billable model is an institutional mechanism to protect BigLaw firm margins, slow automation, and preserve legacy structures.

3. The Role of Trust: a16z’s thesis devotes the most to trust as a structural moat. Trust is important in an industry where reputation and goodwill are at stake. To build trust, startups must focus on establishing a brand among law firms. For example, Harvey has the compounding effect of each additional white-shoe law firm adopting its product, making it more likely that the next law firm will do so under competitive pressure, validation, or FOMO. Battery Ventures does not mention trust but does note that data privacy is a challenge to the adoption of such tools. This is because lawyers may have access to clients' sensitive information, such as documents and records crucial to the client’s case strategy. GC’s thesis on services does not mention trust & liability but focuses more on the economics, long-term ownership, and profitability of such businesses. Furthermore, GC’s thesis highlights that founders need to have access to patient capital, M&A expertise, and workflow management to build a successful AI-powered roll-up.

   
   

   Sequoia’s thesis does not address the role of trust and liability, but it does discuss outsourcing as a wedge strategy and how firms can build trust. Outsourcing usually means the task can be performed externally, a budget line exists, and the buyer is already purchasing an outcome. LegalTech startups should focus on outsourced tasks and expand as the volume of high-volume, intelligent judgment work increases. This basically means that the in-source, judgment-heavy work will expand. The Total Addressable Market (TAM) will expand as firms expand into judgment-intensive work. Lightspeed’s thesis on building implies building collaborative tools that can be slotted into existing workflows with minimal disruption. GenAI tools need a deep understanding of workflows to deliver outcomes, especially given that lawyers are “unforgiving personas”. Tools built have to compound on a day-to-day basis through the self-reinforcement effect and pattern analysis.

   
   

   TLTF’s thesis says little about how LegalTech startups can build trust. However, it merely identifies trust as a fundamental tenet of legal services. This can be inferred from the initial theme of their thesis, which emphasizes that law firms are often critical in disputes, transactions, and transitions. Law firms are trusted by their clients. AI tools will make clients happier by providing value beyond traditional firms.  

   
   

   The Indian legal services industry is a small-knit industry expected to be valued at $2.64 billion by the end of 2026. Lawyers and partners exchange notes at conferences. Trust is essential for LegalTech startups to build, and earning this trust is easier if the founder or the one selling the tool is a lawyer, because there’s a certain way to build trust with lawyers.

   
   

   Without trust, law firms or other legal services customers would not buy products from a LegalTech startup, given that it handles sensitive information. Building trust would allow startups to create a compounding loop where their brand and reputation in legal services improve with every firm that adopts their tool.

   
   

   The core divergence is that LegalTech adoption is influenced not merely by the scale of change the technology brings, but also by institutional trust. In a highly regulated profession like law, trust is an important moat for startups to build on. This applies not only to law but also to healthcare and finance, where small mistakes can cause catastrophic consequences. “Institutional trust” has to be built before building capability.

4. Who is the Customer?: In legal services, there are different profiles of customers that you sell to. Customers in LegalTech include BigLaw firms, mid-sized and Boutique firms, in-house counsel, and clients (AI-native services). GC’s thesis identifies small businesses as its core customers and provides legal services to them. Customer value is created through long-term ownership and tools compounding with every additional matter handled. Sequoia’s thesis identifies clients, law firms, and in-house counsels as the customers. Sequoia’s portfolio companies, such as Crosby (Clients), Harvey (law firms), and Sandstone (In-house counsels), indicate that it has invested in startups with fundamentally different customer profiles. Sequoia’s bets reflect uncertainty about where distribution will accrue across different customers.

   
   

   Battery Ventures does the best of identifying different customer profiles and classifies them as BigLaw, Mid Law, Solo, and In-house counsel, and understands their team size, innovation style, and risk appetite. This provides insight into buyer psychology and how buyers make decisions when buying a LegalTech product. Solo counsels are usually price-sensitive, since they work with minimal associates to cut costs. Value can be created by identifying different buyer personas and building tools according to them. Lightspeed’s thesis is more targeted towards law firms and lawyers. Rather, they are not that interested in startups selling to in-house counsel because, in the context of companies, legal is a cost function and can only be sustainable for LegalTech startups if they expand into other functions, such as Sales, GTM, Finance, and Accounting. The core customer is law firms. Furthermore, Lightspeed is not interested in building a full-stack law firm because it combines the operational complexity of scaling a firm with clients with the business-model complexity that arises from that.

   
   

   a16z’s thesis across customer profiles is to sell to in-house counsel and law firms. However, customer value in this thesis is closely linked to the business models of such customers. The business model of such firms/counsels has to be outcome-based, because a16z believes that pure time-saving will not sell in a business built around the number of hours billed. Customer value to firms or in-house counsel is linked to business models. TLTF’s thesis involved selling a wide range of customers, including law firms, in-house counsels, and clients (services).

   
   

   TLTF is interested in AI-Native Law Firm and refers to it as “Law Firm 2.0.” This firm would have non-lawyer ownership and would adopt an ABS to raise venture capital funding and invest in building AI infrastructure. TLTF’s advantage is its portfolio companies, which handle the operational layer of legal work, such as Firmpilot (a marketing engine for law firms).

   
   

   Indian LegalTech startups are catering to multiple customers, including in-house counsel, law firms, and ALSPs. However, India does not have an AI-native law firm that offers legal services built from the ground up with AI. India is still in the 1^st phase of LegalTech, with BigLaw adopting AI tools. “Autopilot” has not yet achieved traction in India.

   
   

   LegalTech startups have to build a mix of customers, including in-house counsels, law firms, solo practitioners, and clients (for AI-native law firms). Furthermore, only time will tell whether in-house counsel and clients will adopt AI and accept AI-generated work product that has been verified (AI-native law firm).

   
   

   The core divergence is that the legal services market has multiple customer profiles with different price sensitivities, incentives, workflow structures, and adoption patterns. The bets of these firms also point to where they see LegalTech tools accruing value. LegalTech tools involve understanding the buyer persona, as each buyer has different needs, workflows, and domain expertise. The LegalTech market has a diverse customer profile. The winning startup will go deep, focus on a customer profile, and then expand into others.

Axis	GC	Sequoia	a16z	Lightspeed
Delivery Model	Own services	Hybrid	Infrastructure	Augmentation
Billing Assumption	Outcome-based future	Transitional	Hourly persists	Hybrid shift
Trust Strategy	Operational scale	Distribution wedge	Institutional credibility	Workflow integration
Customer	SMB/end-client	Mixed	Firms + in-house	Firms

Conclusion

The venture capital consensus is not merely that AI will improve legal work. LegalTech is reorganizing legal services around price, value, and operations. It involves fundamentally different beliefs about billing rates, customer profiles, trust, and GTM strategy. Based on the divergences and convergences throughout this piece, a framework for evaluating LegalTech startups can be developed for venture capital investors before they decide to invest. The framework consists of:

1. Billing Model: Is your customer billing hourly? Does your customer earn money based on contingencies/outcomes? If they bill hourly, why would they adopt your tool?

2. Type of Task: Where does your core use case fall (intelligence/judgment)? The closer to pure intelligence, the faster the adoption. 

3. Data Moat: What proprietary data set does your startup acquire over time that gives it an edge?

4. Customer Identity: Are you selling to a law firm, in-house counsel, ALSPs, or the end client? Customer Identity is crucial to determining factors such as regulatory risk, sales cycle, and pricing model.

The convergence across these six theses is not accidental. AI capability has crossed the threshold that earlier waves of LegalTech did not reach. Throughout the piece, even in the summaries, we have contextualized all of these theses in India’s context to better understand them.

Furthermore, there’s no Indian-specific LegalTech thesis for firms. A thesis of this nature must include Bar Council restrictions, the MSME access gap, the absence of the contingency-billing model, and differences in billing culture.

For Indian LegalTech, the structural arguments in convergences and divergences, but the regulatory environment, billing culture, and access to justice create a different opportunity, one that deserves its own thesis. LegalTech adoption can be improved by understanding it as an economic and institutional problem surrounding workflows, incentives, trust, and pricing.

---

This article forms part of the LegalTech Investment Thesis Series, a three-part exploration of how venture capital is evaluating AI-driven legal services and LegalTech infrastructure.

Here are the links to Part 1 and Part 2.

Authored by Harshith Viswanath, LegalTech Fellow at the Indian LegalTech Network (ILTN).