| Enterprise vs Team decision per function | Default everyone to Team | Per-function tier choice | Functions handling customer data, regulated work, IP-sensitive material, board-facing material need Enterprise. Others can sit on Team. We help you map functions to tier appropriately. |
| Australian vs US region for data residency | Default to US (cheaper, default) | Per-data-class decision | For most sectors, US region is acceptable. For regulated sectors (financial services, healthcare, government-related), Australian region is the right choice. Trade-off is small price differential, larger compliance clarity. |
| Seat-count budgeting (initial) | Buy seats for everyone who might use it | Start with 50-70% expected users | Buy seats for the functions you are rolling out in the first 60 days. Add seats as additional functions come online. Most mid-market businesses underestimate initial seat count, then over-buy with unused seats. |
| Contract term length | Annual | Annual or 2-year | Annual gives flexibility. 2-year typically gets 5-10% pricing benefit. For organisations with strong AI-future certainty, 2-year is fine. For organisations still proving the use case, annual is the right call. |
| Professional-services hours bundled | Often not bundled | Negotiate 20-50 hours per year | Anthropic professional-services hours bundled into the Enterprise contract can fund pre-rollout setup, custom training, advanced configuration. We help you negotiate the right hours and structure. |
| Custom rate limits | Default tier limits | Negotiated based on use case | For engineering teams using Claude Code heavily, or for batch-processing use cases, custom rate limits are worth negotiating. For most knowledge-worker use cases, default limits are fine. |
| SSO + identity integration | Often deferred | Day-one requirement | SAML SSO via Okta / Azure AD / Google Workspace is part of standard Enterprise. Implementation is straightforward but should be scoped pre-rollout. SCIM auto-provisioning available for larger orgs. |
| AI usage policy + governance framework | Often missing | Mandatory pre-rollout | A 1-page company AI usage policy, signed by CEO and function heads, is non-negotiable before rollout. The policy is the difference between "we use AI" and "we use AI well". We draft it as part of the advisory. |