Benchmark GP: First $1M takes longer now, next $99M happens faster

Chetan Puttagunta breaks down Benchmark's AI investment thesis and what changed in GTM. The firm's seeing legal AI startups hit $100M ARR in 18 months while taking longer to find product-market fit. Direct sales is back, PLG is not the default, and the moat moved from code to distribution.

Benchmark GP: First $1M takes longer now, next $99M happens faster

The Build-Scale Inversion

Benchmark General Partner Chetan Puttagunta says the AI era flipped the startup playbook. Getting to $1M ARR now takes longer than it did in the cloud era. Getting from $1M to $100M happens faster than ever.

Legora, the legal AI company Benchmark backed, spent over a year embedded inside a law firm before launch. Then scaled to $100M ARR in 18 months. That is not an anomaly. It is the new shape: longer research and validation phases, compressed scaling once you ship.

Direct Sales Is the Motion

Puttagunta is skeptical of the PLG-for-everything narrative. The AI companies Benchmark backs are using direct sales and forward-deployed engineers. Why? When code costs approach zero, the defensible work is everything else: customer research, implementation, outcome delivery.

He cites examples collapsing 180-day sales cycles into 30 days through tight pilots and magical demos. But those pilots require sales teams who understand the customer's workflow, not just the product.

What Benchmark Actually Looks For

The firm makes maximum 10 investments per year. In June 2026, they raised their first growth fund: $2 billion across two vehicles, including $1.25 billion for later-stage deals. That is a shift from their traditional sub-$500M early-stage focus.

Puttagunta's investment criteria: technical insight that creates demand pull. Not just a better feature. A founder who spent time in the customer's world and saw what was broken. Distribution advantage matters more than product defensibility when the product is code and code is cheap.

Implications for Sales Teams

If you are selling AI products, expect longer pilots and more technical validation. If you are building sales teams at AI startups, plan for direct sales headcount and forward-deployed engineers, not just SDRs feeding product-led funnels.

The value moved. It is not in writing the code anymore. It is in understanding the customer well enough to deliver an outcome they will pay for. That requires salespeople who can operate more like consultants and engineers who can operate more like account managers.

Benchmark's portfolio companies are hiring for that hybrid profile. Worth watching if you are planning your next role or your next hire.