The Claude team rollout: from one believer to a fluent shop.
You're convinced; now there's a team. The rollout playbook is tool-agnostic at heart — policy, recipes, champions, rhythm — with Claude-specific choices about plans, seats, and which strengths to lead with.
01 Plans and seats, decided by role
Claude's team offerings give you central billing, higher limits, and workspace administration — the governance wrapper individual accounts lack. Seat strategy beats blanket buying: lead with the reading-and-writing-heavy roles — whoever drafts proposals, reviews contracts, answers hard emails, produces reports. That's where Claude's signature strengths pay fastest and where your proof-of-value stories will come from. Expand on evidence, not faith (the costs lesson keeps the per-seat math current).
02 The one-page policy (borrow ours)
Green/yellow/red plus the blame rule — the same format from our ChatGPT rollout lesson transfers wholesale; adapt it in twenty minutes. Add the two connector lines from the security lesson if anyone will touch MCP, Computer Use, or Cowork. One page, shipped this week, beats a perfect policy shipped never.
03 Seed recipes that match Claude's strengths
Three recipes per role, demonstrated live on real work — never "here's Claude, explore."
Same physics as every tool rollout: novelty week, then quiet abandonment among those without a personal win. The countermeasures are boring and they work: a named champion who isn't you, a monthly 20-minute show-and-tell of two real wins, and one measured number (hours saved, drafts shipped, turnaround time). What gets counted keeps getting used.
04 The multi-tool reality
Your shop may run Claude alongside ChatGPT or Copilot — that's normal and usually right (see the honest comparison). The rollout implication: assign each tool a LANE in your recipes ("contracts and proposals → Claude; scheduling and email triage → Copilot") instead of letting every person relitigate the choice daily. Lanes reduce confusion, and confusion — not capability — is what kills adoption.
Draft the policy (adapt ours), pick the first two seats by role, and write the three recipes for seat one. That's a rollout plan — most companies never get this far.
Open Claude →This week's challenge
Two weeks: seats live for the two writing-heaviest roles, policy shipped, three recipes each demonstrated on real work, champion named, verdict meeting on the calendar with one measured number. Renew or rethink on evidence. (Want it done alongside you? That's our advisory practice.)