Agent-Native Go-To-Market
Definition
The product positioning thesis that the Brain must be designed for agents as the primary consumer of its context — not humans browsing a wiki — because within 1–2 years most knowledge work will be done by agents on behalf of humans. Articulated most clearly by saar-arbel in 2026-05-06-brain-os-strategy-brainstorm.
Key points
- The bet: “anyone who is not working primarily through agents in 1–2 years will look like they are working in 1995.” The Brain must therefore be queryable by agents from day 1, not retrofitted.
- Companion framing, not replacement. The product is not “do your work for you” — it is “make your agents better at doing your work.” This dodges the worst objections (job-replacement panic, productivity-tool commoditization).
- Adoption signal: non-engineers at monday (Enablement, growth) are already using cursor + NoteTaker + Brain daily — see nizan-shifman’s sister anecdote. The agent-native pattern is leaking out of engineering faster than anticipated.
- First-mover matters but is not sufficient. The pitch must explain why the Brain wins agent-share over glean (search-tool tier), sierra (FDE-heavy enterprise), wonderful (adjacent), cursor/Claude/ChatGPT (eating from below).
- Implication for product surface area: less Obsidian-style human browsing UX, more MCP / API / structured-context surfaces; permissions and authorization become first-class because agents will be the ones traversing the brain.
Evidence
- 2026-05-06-brain-os-strategy-brainstorm — Saar’s central argument; Nizan’s sister-on-Enablement anecdote as evidence of the spread beyond engineers.
- 2026-05-27-directions-vertical-pivot-and-prediction-markets — Saar’s “everyone is approaching the agentic-org horizontal” observation; the agent-native thesis is no longer a differentiator on its own and must combine with vertical use-case ownership (vertical-use-case-led-brain).
- 2026-05-01-entree-capital-enterprise-ai-spend-map — external market data: agentic AI is named as its own emerging category, overtaking chatbot spend in 2026–27 and reaching ~70B ARR by 2028, with Claude Code (>50% enterprise) as the breakout line — directly relevant to the cursor-style agent harness the team rides daily.
- 2026-05-16-alonhuri-linkedin-ai-native-growth-hacking — extends the agent-native pattern from context-consumption (the team’s framing) to GTM/growth-loop generation: alon-huri argues AI has solved every classical startup bottleneck except customer acquisition (CAC up 222% in 8 years; 1 ARR), and that growth hacking must shift from a single human artisan running serial hacks to an autonomous closed-loop “growth lab” with agents at every node (research → judgment → thesis → execution → measurement → learning, ~50 hacks in parallel) and humans only at strategic-judgment seams. “For founders building today, this isn’t a feature you add later — it has to be the GTM core from day one.” Direct implication for the Brain: the research agent’s required context corpus is exactly what the Brain produces. Skepticism caveat — Alon is Managing Partner at team8; treat as advocacy.
- 2026-01-09-alonhuri-linkedin-vibe-acting — the third layer of agent-native pattern: agents that take real-world actions, not just consume context (team’s framing) or generate growth experiments (Alon’s earlier post). “In the last decade we built systems that wait for us to click. In the next decade we’ll build systems that understand what we want and act.” Argues the LLM giants have a structural ceiling — they can’t access bank accounts, medical records, legal docs, business systems, deep personal context — and that is where vertical applications + Vibe-Acting infrastructure win. Direct implication for the Brain: it must hold the data the action-agents need to act safely. Skepticism caveat — Team8 portfolio alignment.
Update (2026-05-27)
The “make your agents better” framing remains, but the team’s center of gravity has shifted: rather than sell agent-native context to whoever wants it, own the agents in a chosen vertical and let agent-native context be the moat. Agent-native is now necessary-but-not-sufficient. See vertical-use-case-led-brain for the full reframe.
Sobering observation from the meeting: outside coding, very few production agents have changed anyone’s life yet. saar-arbel’s framing — “everyone wants the adoption story; nobody has it.” Implication for this topic: the GTM bet on agent ubiquity must be paired with picking workflows where the agents have already proven worth paying for.
Open questions
- Agent-to-agent workflows — if every employee has their own agent, does the Brain become the substrate for agents talking to each other rather than humans talking to each other? What does that interface look like?
- How does the Brain stay compatible across agent ecosystems (Claude managed agents, Cursor’s in-progress equivalent, OpenAI’s forward-deployed PM model, future entrants)?
- Can “agent-native” be the differentiator vs glean, or will Glean ship the same surfaces by the time the Brain is enterprise-ready?
- Which currently-deployed agent use cases (e.g. monday-agents-month entries like Engy, Simona, Amit’s mobile-port) actually compound with deeper Brain context vs being adequate as thin wrappers?