Your First Week With Salestrics AI: A Launch Playbook for Revenue Teams
Day zero checklist (before Monday)
If you are rolling this out to a team, confirm these before day one:
- Buyer mail synced and attached to opportunities — not only in personal inboxes
- Pipeline stages and owners updated on active deals
- Admin access to AI Guardrails for whoever owns org AI Instructions
- One written sentence: the workflow you are trying to fix this week
Skipping this step is the most common reason week-one adoption feels like “another chatbot tab.”
Pick one workflow first
Before Monday, write down one ritual you want AI to shorten. Examples:
- Sunday night pipeline exports before standup
- Follow-up drafts after demos
- “What happened on this account?” Slack threads
Everything below orbits that sentence. If you cannot name the workflow, you cannot measure whether week one worked.
What each role does in week one
| Role | Week-one job |
|---|---|
| Founder / admin | Write org AI Instructions; set default model; define the one workflow to fix |
| RevOps | Audit mail-on-deals and stage hygiene; fix before blaming Briefings |
| Sales | Use /email after demos; /pipeline before standup |
| Support | Ask Assistant account questions that used to require CRM + Resolve tab switching |
Everyone does not need every surface on day one. Match access to the workflow you picked on day zero.
Day one: admin setup
Spend thirty minutes in Admin → AI Guardrails before inviting sellers:
- Company context — what you sell and who you sell to
- Communication style — direct vs formal; whether drafts mention pricing
- Terminology — product names, acronyms, competitor language to avoid
- Restrictions — no SLAs, discounts, or roadmap dates in drafts
- Default model — v2 for investigations; v1 for fast everyday drafts
Permission defaults (seed-stage)
| Role | Day-one access |
|---|---|
| Founder / admin | Full AI surfaces + guardrails + MCP if needed |
| Sales | Assistant, Compose, Briefings read |
| Support | Assistant + Resolve context |
| Finance | Analyze; Ledger KPIs per finance permissions |
Day two: Morning Brief
Ask leaders to read Morning Brief on the dashboard before standup. One success signal: someone references a deal or thread they would have missed without it.
If Briefings cite wrong records, fix CRM hygiene — stages, owners, mail on opportunities. Bad data produces bad briefs regardless of model quality.
More on Briefings: proactive daily intelligence.
Day three: three slash commands
Teach these in a 15-minute huddle — not the full list:
/pipeline— before standup/email— after demos; drafts from the thread on the opportunity/timeline— before customer calls
For one week, route buyer follow-ups through /email instead of pasting
threads into an external chatbot. If Assistant needs paste to answer, log it — that is
a data or permissions gap, not a reason to open another tab.
Mail-specific tips: AI email for sales teams.
Day four: MCP (optional)
Skip this if day three did not stick. When you are ready, MCP integrations let Assistant reach Stripe, Slack, Sentry, Notion, Linear, or warehouse Postgres when the answer is not on the CRM graph.
Enable one integration that matches a recurring argument:
- Stripe + CRM — deal stage vs subscription state
- Slack + CRM — internal thread context on accounts
- Notion + pipeline — mutual action plans beside opportunities
Document which questions it answers. Require Planning approach visibility before tool calls touch billing data.
Day five: retro
Twenty minutes. Five questions:
- Did Morning Brief change what we discussed in standup?
- How many follow-ups used
/email? - Could a second seller answer an account question without pinging the founder?
- Where did we still paste into external AI — and why?
- What org AI Instruction would have prevented a bad draft?
Mostly “no” answers usually mean mail not on deals, stale stages, or sellers blocked from Mail AI — fix the graph before swapping models.
Common mistakes
- Skipping org AI Instructions — generic drafts kill trust fast
- Enabling every MCP integration on day one
- Using v2 for short drafts when v1 is faster
- No written success metric from day zero
Reference
Launch inventory: what shipped on OpenAI Day · Changelog #9 · demo at ai.salestrics.com · vs HubSpot AI.