New Rep Territory Launch Chain: ICP → Top 10 Accounts → Outreach

What this chain helps you accomplish

Turn a new territory assignment into a focused first-30-days prospecting plan rather than guessing which companies to pursue and sending generic emails.

Six connected prompts help a new sales rep build a Territory ICP & Scoring Guide, establish a verified Candidate Account Shortlist, produce Account Intelligence Briefs for up to ten target companies, rank them in an Account Priority & Buying-Role Map, create Three-Touch Outreach Drafts for appropriate contacts, and assemble a 30-Day Territory Launch Plan with evidence, ownership and review checkpoints.

The account work is designed to be repeated, one real company at a time. Unknown account information stays UNKNOWN; industry patterns remain hypotheses, not established buyer pain; named prospects must be verified. The plan can finish with fewer than ten accounts when the territory data do not support ten credible selections.

What you receive: clear seller instructions and copyable AI prompts to run in ChatGPT, Claude, Gemini or another approved text-capable tool. Promptifi does not independently search your CRM, find contacts, send messages, run sequences, book meetings or produce sales outcomes.

When to use it

• New territory or new rep: An SDR, BDR, AE or account manager needs to understand who is a realistic fit, where to focus and how to begin outreach in the first month.
• Territory reassignment or vertical expansion: A seller inherits a list of accounts and must distinguish high-fit targets from accounts that require research or are out of scope.
• Manager-led onboarding: A frontline manager wants a repeatable, reviewable process for ICP assumptions, ten target accounts, buyer-role choices and first messaging.
• Before launching an outbound motion: A rep needs account-specific evidence and truthful first-touch sequences that can be checked against sales policies before any message is sent.

Not for: Mass-generating fictional accounts or prospect identities, circumventing territory ownership, relying on outdated customer examples, or treating an AI-predicted pain point as a buyer admission. This resource does not replace product enablement, company prospecting policy or actual discovery.

The first 30 days are a planning horizon, not a promise that ten accounts can be researched, contacted or converted within a fixed time. Sellers can complete only the accounts for which appropriate sources, contacts and permissions exist.

Before you begin

Required tool: An organization-approved AI assistant that accepts text. No paid plan, live research, CRM connector, ZoomInfo, email sender or automation is necessary to run the prompts. Web search and CRM data may improve evidence when available and authorized, but the AI must never claim to have read a tool or webpage it cannot actually access.

Required territory inputs: Gather the information below before beginning. Use the same placeholders in the Starting Territory Context. Missing optional information can be marked UNKNOWN or NONE; do not fabricate it.

• {{TERRITORY_SCOPE}} — Required. Geographies, industry/vertical, company segment, named-account limits, existing account ownership and exclusions from the manager or territory assignment. Example: Midwest U.S. midmarket industrial distributors excluding house accounts. If no rules exist, obtain the territory boundary before selecting accounts.
• {{OUR_OFFER_AND_LIMITS}} — Required. Actual product/service, best-supported use cases, delivery prerequisites, known limitations and approved positioning. Source: product/enablement team. Example: platform for managing inbound distributor quote requests; integrations require scoping. If unverified, do not claim a capability.
• {{BEST_FIT_CUSTOMER_EXAMPLES}} — Recommended; may be NONE. Two or three manager-approved anonymized customer patterns: industry, size, use case, reason for value and observed outcome (only if substantiated). Example: regional distributor with several branches used a centralized quote-routing capability. If no examples exist, label the resulting ICP a working hypothesis.
• {{POOR_FIT_CUSTOMER_EXAMPLES}} — Recommended; may be NONE. One or two anonymized poor-fit or disqualified patterns with evidenced reason. Example: very small single-location organizations with no workflow complexity. If absent, use only known hard product/territory exclusions.
• {{APPROVED_PROOF_AND_CLAIMS}} — Optional. Shareable product or customer claims, case studies and their approval/limitation. Example: approved feature sheet, Oct. 2026, supports assigning request ownership; no quantified ROI approved. If NONE, do not invent proof.
• {{CANDIDATE_ACCOUNT_LIST}} — Required as a field; may initially contain fewer than ten. Real companies from manager-approved territory list, CRM or authorized research, one per line, with website/domain if verified, existing owner/status and C# identifier if known. Example: C01 — Cedarfield Parts — cedarfield.example (fictional illustration). If fewer than ten, the AI must not create fictional filler; research more with an enabled verified source or report the gap.
• {{ACCOUNT_SOURCE_PACK}} — Optional initially; required for substantiated scoring. Per-company research excerpts labeled A01-S1, A01-S2, etc. with publisher, date, verified URL if available, and exact supported claim. Example: A01-S1 — company newsroom — 2026-10-15 — announced a new branch. A link without retrieval is not evidence.
• {{CONTACT_AND_ROLE_SOURCES}} — Optional. Verified professional names/titles or role-only targets from company website, approved CRM or permitted research, tied to account and source/date. Example: no named operations VP verified; role target only. Never generate a plausible contact name or email address.
• {{RESEARCH_ACCESS_AND_RULES}} — Required. State whether live web search is enabled, whether CRM data was actually supplied or connected, which sources are allowed, and whether the AI may browse. Example: web access available but CRM not connected; rep must validate URLs. If NO WEB ACCESS, use only pasted sources and identify research gaps.
• {{SENDER_AND_OUTREACH_RULES}} — Required before any external use. Company-approved sender, jurisdictions, contact/opt-out policy, suppression check, channel restrictions, approved templates and ownership rules. Example: sender verified through approved email system; prospect suppression check pending. If unknown, label all email content DRAFT ONLY / NOT SEND-READY.
• {{MANAGER_GOALS_AND_CAPACITY}} — Optional. Approved first-month priorities, training requirements, expected research/review capacity, number of permitted contacts and reporting cadence. Example: manager review in Week 1; research first ten accounts in Week 2. If absent, use a suggested plan without invented quotas or approvals.

Keep three evidence layers separate: T# = company-wide territory/ICP evidence; C01–C10 = candidate account identifiers (assigned consistently by the chain); A01-S1 = source 1 for account C01; P# = approved product/seller proof. Record source/publisher, date, URL if actually verified, and the exact claim supported. If a buyer has not said they experience a problem, call it an OUTREACH HYPOTHESIS, not confirmed pain.

Scoring approach: Five fit criteria, up to 20 points each (maximum 100). Count only evidenced criteria; show coverage/missingness separately. Do not award a fabricated score based on broad industry stereotypes. Fit, current timing signal and actual buying intent are three different things.

Data and outreach safety: Use authorized AI tools and approved business information. Do not paste confidential customer data or sensitive personal information unless permitted. For commercial email, follow applicable jurisdiction and company policies. In the U.S., CAN-SPAM applies to commercial B2B email and requires truthful sender/subject information, a valid postal address and working opt-out processes; never invent these. A draft email is not consent to send.

Repeat-use note: Researching ten genuine accounts and reviewing contacts can take longer than the AI text generation. Do not rely on unsupported time-savings claims.

How to run the chain

1. Open a new approved AI conversation. Fill and paste the Starting Territory Context from the Chain Steps. Include your actual offer, territory boundaries, customer-pattern evidence, candidate list, permitted sources and manager constraints.
2. Run Prompt 1 once to build your Territory ICP & Scoring Guide. Check all five 20-point criteria, exclusions and proof sources. Approve the corrected version explicitly.
3. Run Prompt 2 once to validate and deduplicate candidate accounts. It creates a provisional shortlist (up to ten, or fewer with a documented gap), without inventing companies, domains or territory assignment. Approve the shortlist.
4. Run Prompt 3 once per shortlisted account, supplying the real company's approved sources each time. Save each named Account Intelligence Brief with C# ID and version. Stop after each company to check dates, identities, source links, assumed pain and scoring evidence. Do not claim ten researched briefs until ten actually exist.
5. When the briefs have been reviewed, run Prompt 4 to score, rank and map buyer roles for accounts with sufficient evidence. Review fit-versus-timing signals and whether contacts are actually verified. Approve the Account Priority & Buying-Role Map.
6. Run Prompt 5 once per selected account with its approved Account Intelligence Brief and the approved ranking/map. Create and review the three-touch draft. Where the named contact or trigger is unknown, keep only honest role-based drafts and mark NOT SEND-READY.
7. After your completed outreach drafts exist, run Prompt 6 to compile the first-30-days territory plan, coverage dashboard, account tracker, quality checklist and manager review agenda. Do not represent drafted messages as sent, outcomes as won, or unverified accounts as researched.

Approval protocol: Write APPROVED — [Exact Artifact Name] — v1. Use only this version as the input to later prompts. For account-specific artifacts include C# and account name. A corrected draft needs a new version and explicit seller approval; it does not automatically replace a previously approved source.

Keep the chain in one conversation where possible. If your AI context becomes too long, start a new chat with the complete Starting Territory Context, approved ICP/shortlist, relevant C# briefs, source register, approvals, and next prompt. An optional handoff template is provided at the end of Chain Steps.

Stop conditions: If there are no legitimate candidate accounts, stop at account sourcing; if evidence is insufficient, mark research required instead of assigning a false high score; if a contact or compliance condition is unverified, drafts are not send-ready. A suggested Day 1/5/10 sequence is illustrative—not an automated schedule or guarantee.

Done means: A usable ICP, a real account shortlist of whatever verified size exists (target ten), individual sourced briefs, defensible priority tiers, reviewable outreach drafts and a practical first-month operating plan. Any real sending, prospect enrichment, CRM updating or scheduling requires separate authorization and appropriate tools.