Geographic Expansion And Operational Scaling In The Us
Key takeaways
- Legora went from zero US headcount at the start of the year to 50 people, is opening a Manhattan office targeted to reach 150 people, and plans three additional US offices this year.
- Max disputes that Harvey is the default first name in legal AI anymore and disputes the narrative that Harvey 'won the US' while Legora 'won Europe'.
- After raising $10M from Benchmark and then $25M from Redpoint about a month later, Max told the board Legora would not sell the product for the next six months; the pause was used to rebuild infrastructure for reliability/scalability, with a claim that by Oct 1, 2024 it could onboard 1,000 lawyers per day comfortably, and that continuing to sell without fixing scalability/reliability would have caused churn.
- Max expects the top model for Legora’s work in the next 24 months to be either Anthropic Claude or Google Gemini, with OpenAI declining in relevance for their enterprise use case, and he perceives Anthropic moving toward enterprise needs while OpenAI moves toward B2C.
- Legora currently charges on a per-seat basis, which Max says is optimal for the buyer but not optimal for Legora due to potentially unsustainable LLM costs from heavy users, and he expects a switch to consumption-based pricing within about three years when customers are ready; he also believes legal software will shift to consumption pricing later than other enterprise tools.
Sections
Geographic Expansion And Operational Scaling In The Us
The corpus provides a concrete US expansion footprint and a gating heuristic that relied on reference wins before building local presence. It also introduces an operational explanation for faster scaling in the US labor market, tied to termination periods.
- Legora went from zero US headcount at the start of the year to 50 people, is opening a Manhattan office targeted to reach 150 people, and plans three additional US offices this year.
- The US has become Legora’s largest market by revenue, and Max expects US revenue to exceed Europe by the end of Q1.
- Legora’s go-to-US heuristic was to first sign and successfully serve two AmLaw 200 firms from Europe before opening the US market, citing Cleary Gottlieb and Goodwin Procter as reference wins.
- Max argues the short US employment termination period (about two weeks) versus Europe (around three months) is a structural advantage for rapidly scaling a team.
Market Structure And Competitive Narratives In Legal Ai
The corpus asserts a concentrated market outcome (winner-take-all) while also emphasizing that customer loyalty is currently low due to short contracts and extended-pilot behavior. It also includes an explicit challenge to popular leadership/geography narratives, implying that perceived incumbency may be less stable than commonly framed in this category.
- Max disputes that Harvey is the default first name in legal AI anymore and disputes the narrative that Harvey 'won the US' while Legora 'won Europe'.
- Customers are not loyal in legal AI right now because firms treat deployments as extended pilots and 'call options' on AI, typically signing one-to-three-year contracts rather than five-year terms.
- Legal AI platforms will be winner-take-all, with the top provider capturing roughly 90% of the market and ranks two through ten splitting the remaining ~10%.
Enterprise Adoption Bottlenecks: Implementation, Activation, Reliability
A repeated constraint is that workflow-transforming deployments require manual activation and reliability work, unlike simpler point solutions. The reported decision to pause sales to rebuild infrastructure is presented as necessary to avoid churn, with throughput onboarding capability used as an operational readiness benchmark.
- After raising $10M from Benchmark and then $25M from Redpoint about a month later, Max told the board Legora would not sell the product for the next six months; the pause was used to rebuild infrastructure for reliability/scalability, with a claim that by Oct 1, 2024 it could onboard 1,000 lawyers per day comfortably, and that continuing to sell without fixing scalability/reliability would have caused churn.
- Enterprise deployments that change how people work require significant upfront manual implementation and activation effort, and Legora uses ex-top-tier-law-firm legal engineers who are forward-deployed to ensure customer success.
- Large law firms run vendor bake-offs and select an AI partner based on both immediate outcomes and belief in the vendor’s long-term vision for AI-enabled legal work.
Model-Provider Switching And Enterprise Orientation
A concrete provider mix change (from only one provider to majority another) is attributed to suitability and prompting dynamics, and is paired with a forecast of future provider leadership. A separate expectation suggests the frontier constraint may shift toward inference economics rather than capability for many tasks.
- Max expects the top model for Legora’s work in the next 24 months to be either Anthropic Claude or Google Gemini, with OpenAI declining in relevance for their enterprise use case, and he perceives Anthropic moving toward enterprise needs while OpenAI moves toward B2C.
- Legora switched from being only OpenAI (2023 through most of 2024) to being majority Anthropic around the Sonnet 3/3.5 era due to perceived model suitability and prompting dynamics.
- Frontier models are approaching a level where, for many tasks, focus should shift from capability improvements to cost optimization.
Unit Economics And Pricing Model Transition
Seat-based pricing is described as buyer-friendly but potentially misaligned with variable model costs, leading to an expected shift toward consumption-based pricing on a multi-year horizon. Current margins are characterized as acceptable but deprioritized during a land-grab phase, indicating that pricing and cost controls are treated as later-stage optimizations.
- Legora currently charges on a per-seat basis, which Max says is optimal for the buyer but not optimal for Legora due to potentially unsustainable LLM costs from heavy users, and he expects a switch to consumption-based pricing within about three years when customers are ready; he also believes legal software will shift to consumption pricing later than other enterprise tools.
- Max states Legora’s margins are only 'okay' today and the current phase is a land-grab rather than margin optimization.
- Legora expects to move away from seat-based pricing to consumption-based pricing likely within about three years, timed to when customers are ready to buy on consumption.
Watchlist
- Max expects the top model for Legora’s work in the next 24 months to be either Anthropic Claude or Google Gemini, with OpenAI declining in relevance for their enterprise use case, and he perceives Anthropic moving toward enterprise needs while OpenAI moves toward B2C.
- AI-driven displacement of knowledge work may appear in labor statistics within the next 12 to 24 months.
Unknowns
- What is the breakdown of the reported $7M single-day ARR event by number of customers, net-new vs expansion, and contract duration/terms?
- How many of the reported 750 customer firms are firmwide deployments versus limited pilots, and what are typical seat counts per deployment?
- What are cohort-based logo retention and net revenue retention numbers after 12–24 months, especially for early 2024 and 2025 cohorts?
- What are Legora’s current gross margins and LLM cost drivers per user/matter, and what usage patterns cause the reported seat-pricing misalignment?
- What specific procurement or budgeting constraints delay consumption pricing adoption in legal compared to other enterprise categories?