Dorfman AI-Native Sales Org Build

// When should you use the Dorfman AI-Native Sales Org Build?

Use this skill when demand has outpaced your ability to hire and onboard sales reps, or when you need to dramatically increase AE productivity and coverage without sacrificing bar, culture, or customer experience. Especially relevant when you already have a working sales tech stack and want to make it compound rather than replace it.

// What inputs do you need before building an AI-native sales org?

• Current tech stack inventoryrequired
  List every tool currently in use across sales, deal desk, legal, RevOps, customer support, and billing — including any that already have Claude or AI features embedded.
• Four constraints maprequired
  Define your four constraints: (1) demand volume you cannot staff for, (2) AI/Claude capabilities already inside your stack, (3) cross-functional dependencies (legal, deal desk, RevOps, billing, CS), (4) headcount growth ceiling and culture bar you refuse to drop.
• Top-rep behaviour auditrequired
  Document the specific activities, research habits, follow-up patterns, and collateral practices of your highest-performing reps — this becomes the raw material for Skills.
• Lead qualification criteriarequired
  The signals and firmographic/behavioural data that distinguish enterprise sales-funnel leads from self-serve candidates.
• Support function ticket taxonomy
  The categories of requests AEs send to deal desk, legal, RevOps, billing, and compliance, plus the policy and precedent documents that govern resolutions.

// What are the core principles of the Dorfman AI-Native Sales Org Build?

// How do you apply the Dorfman AI-Native Sales Org Build step by step?

1. 1

   Map the Four Constraints

   Before touching any tooling, explicitly name your four immovable constraints: (1) demand you cannot staff for, (2) Claude already embedded in your existing stack, (3) cross-functional functions that must scale alongside sales, (4) headcount ceiling and culture bar. Every subsequent decision is made inside these constraints. Do not skip this step — building without constraint clarity produces seven isolated tools, not a cohesive system.

2. 2

   Audit and commit to your existing stack

   List every tool in use. Identify which already have Claude or AI natively threaded. Mark each tool's role in the lead journey (enrichment, routing, system of record, call intelligence, contracts, deal coordination, support). Decide which six tools form the core architecture of a lead's journey from inbound to closed-won. Do not add new tools yet — the discipline is to build coherence across what you have.

3. 3

   Design the Two-Funnel qualification architecture

   Use Clay and Claude to enrich, evaluate, and qualify every inbound lead. Build two parallel routing paths: (a) Self-Serve Funnel — Claude and an AI-guided support tool (e.g. an Intercom Fin-type product) handles the full buying journey through to provisioning, billing, onboarding enrolment, and terms acceptance, generating real ACV with no human AE required; (b) Sales Funnel — lead goes to BDR for qualification and AE routing. Define the qualification signals that determine funnel assignment. Launch the self-serve funnel as an MVP first, iterate into production. Track the percentage of new enterprise logos coming through self-serve as your primary health metric for this funnel.

4. 4

   Thread Claude through each stage of the six-tool lead journey

   For each stage in your core six-tool architecture, identify what Claude does that the tools cannot do alone: account research and prioritisation at lead creation; pulling historical context from Slack, docs, and call recordings into the account record before AE assignment; proposal generation that checks policy, aligns to customer history, and auto-uploads to contract tooling; forecast submission that reconciles data across systems so manager review is discussion-oriented, not data-gathering. Claude should update records, surface context, and generate drafts — humans inspect, decide, and approve.

5. 5

   Make Slack (or Teams) the front door for all support functions

   Eliminate the sea of DMs and approval-chasing. Instrument a single Slack-in, ticket-out workflow for every support function: deal desk, legal, RevOps, billing, compliance, customer support. Claude triages every incoming ticket and either (a) resolves it autonomously based on precedent and policy, or (b) escalates it to the right human with all relevant context pre-pulled from email, CRM, call recordings, and Salesforce — and notifies the AE so they can set customer expectations. One step per closed-won checklist item ensures governance without manual coordination.

6. 6

   Conduct the top-rep behaviour audit and encode findings as Skills

   Interview and observe your highest-performing reps. Document their specific pre-call research process, follow-up discipline, competitive positioning habits, and collateral creation approaches. These become the raw inputs for Skills. A Skill is a combination of MCP connectors and Claude prompts that replicates the best-rep behaviour at scale. Build the five core Sales Skills: (1) Morning Brief, (2) Call Prep, (3) Customer Follow-Up, (4) Competitive Intel, (5) Create an Asset. Package them as a Sales Plug-in issued to every rep at onboarding.

7. 7

   Build and deploy the five core Sales Skills

   Morning Brief: single daily prompt pulling from calendar, email, Slack, CRM, call recordings, marketing events, and centralized initiatives — delivered to Slack or inbox at a fixed time each morning with three prioritised actions. Call Prep: invoked via shortcut before each call; outputs stakeholder research, historical context, ideal call outcome, discovery questions, competitive landscape, partner ecosystem, and public signals about customer needs. Customer Follow-Up: extracts action items from email, calls, CRM notes, and Slack; drafts responses; deposits them in the rep's email provider; sends a summary; flags outstanding items in next morning's brief. Competitive Intel: generates dynamic, interactive battle cards tailored to the specific customer and deal, updated in real time — not static quarterly product marketing decks. Create an Asset: given deal stage, stakeholder profile, and customer needs, generates custom collateral (one-pager, proposal, ROI calculator, interactive HTML, prototype) on demand, on brand, during or between calls. Skills must encode brand standards so output is never generic.

8. 8

   Instrument dynamic coaching at the manager layer

   Do not rely on a static methodology for rep coaching. Surface six coaching moments per week per rep using Claude, calibrated dynamically to current business priorities — what mattered last month may not be the top priority this week. Coaching moments should reflect the live competitive environment, product changes, and deal patterns. Forecast calls are reserved for discussion and manager/AE support — not for data reconciliation, which Claude handles in advance.

9. 9

   Establish the AGI Pills growth loop

   Set the team norm: every rep's job is to identify one thing each day that can be done manually once and then encoded so Claude does it next time. Frame this as a growth mindset — nobody has done this before, and incremental improvement each day compounds. Track which Skills are being used, where AEs are still doing manual work, and treat those gaps as the next encoding candidates. The org gets smarter as Claude accumulates context across every deal.

// What does the Dorfman AI-Native Sales Org Build look like in practice?

// What mistakes should you avoid when building an AI-native sales org?

• Bolting Claude on as a seventh standalone tool rather than using it as connective tissue between existing tools — this produces six siloed experiences, not a cohesive customer journey.
• Treating self-serve as a downgrade or a lesser tier. Self-serve is a fully valid, revenue-generating funnel that removes friction and serves a real buyer segment — not a consolation path for prospects who couldn't get a human.
• Replacing the existing tech stack instead of building on it. Years of investment, customisation, and institutional memory are in your current tools. The discipline is coherence and threading, not replacement.
• Scaling only the AE layer without giving the same elasticity to deal desk, legal, RevOps, billing, and customer support. Sales is not an island — if supporting functions don't scale, they become the new bottleneck.
• Using a static coaching methodology when the business environment is changing weekly. Coaching moments must be dynamic, recalibrated to current priorities, competitive shifts, and product changes — not set quarterly.
• Shipping AI-generated collateral that is generic or off-brand ('AI slop'). Every asset generated through the Create an Asset Skill must be on-brand, tailored to the specific deal and stakeholder, and designed to help win — encode brand standards into the Skill.
• Allowing forecast calls to remain data-gathering sessions. Claude should handle all data reconciliation in advance so that forecast calls are reserved exclusively for discussion, deal strategy, and identifying where AEs and managers need help.
• Sacrificing recruiting bar or culture in order to move faster on headcount. If you can't hire at your standard, build systems to amplify the people you have rather than diluting quality.
• Assuming Skills are a one-time build. Skills must evolve as the business evolves — what was the top coaching priority last month may be irrelevant this month. Treat Skills as living artefacts.

// What are the key terms in the Dorfman AI-Native Sales Org Build?

AI-Native Sales Org

A sales organisation in which Claude and AI tooling are the connective tissue of the entire go-to-market system — not add-ons or features — enabling the team to absorb demand and operate at a productivity level disproportionate to headcount.

Skills (Stills)

Encoded best practices of top-performing reps, built as combinations of MCP connectors and Claude prompts, that any rep can invoke to replicate expert behaviour on demand. The five core Sales Skills are: Morning Brief, Call Prep, Customer Follow-Up, Competitive Intel, and Create an Asset.

Sales Plug-in

A packaged bundle of MCP connectors and Skills issued to every rep at onboarding. It contains all five core Skills and gives the rep immediate access to best-rep behaviours from day one, compressing ramp time.

Morning Brief

A daily AI-generated prioritisation summary delivered to each rep's Slack or inbox at a fixed time, pulling from calendar, email, CRM, call recordings, Slack, and marketing systems to surface the three most important actions for the day.

Call Prep

A pre-call briefing Skill that outputs stakeholder research, historical deal context, ideal call outcome, discovery questions, competitive positioning, partner ecosystem signals, and public customer statements — delivered as a one-pager before every call.

Customer Follow-Up Skill

A Skill that extracts all action items from email, call recordings, CRM notes, and Slack; drafts follow-up responses; deposits them in the rep's email client; and flags outstanding items in the next morning's brief — enforcing a 24-hour follow-up SLA.

Competitive Intel Skill

A dynamic, interactive battle card generator that produces a tailored, real-time competitive matrix for the specific customer and deal in context — replacing static quarterly product marketing battle cards.

Create an Asset Skill

A Skill that generates completely custom, on-brand collateral (one-pagers, proposals, ROI calculators, interactive HTML files, prototypes) tailored to the specific deal, stage, and stakeholder — accessible to every AE for every deal, not just top-tier accounts.

Two-Funnel Architecture

The parallel operation of a Self-Serve Funnel (AI-qualified leads guided to enterprise plan purchase and onboarding with no human AE) and a Sales Funnel (AI-qualified leads routed to BDR and AE). Both funnels are considered equally legitimate revenue paths.

Connective Tissue

The role Claude plays across the existing tech stack — not a standalone tool, but the layer that makes all existing tools communicate, share context, and create a seamless customer and AE experience.

Slack-In, Ticket-Out

The design pattern where Slack (or Teams) is the single front door for all support function requests. Reps submit via Slack, Claude auto-generates and triages the ticket, resolves or escalates with full context pre-assembled, eliminating the sea of DMs and approval-chasing.

AGI Pills

The growth mindset norm for the AI-native sales team: do one thing manually once, then encode it so Claude does it every subsequent time. The goal is for Claude to spot patterns before humans can, getting incrementally better at the sales motion each day.

Four Constraints

The four immovable parameters that define the build: (1) demand you cannot staff for, (2) Claude already embedded in existing stack, (3) cross-functional functions that must scale alongside sales, (4) headcount ceiling and culture bar you refuse to compromise.

Closed-Won Governance

A one-Skill-per-step checklist Claude works through at deal close to ensure every cross-functional requirement — provisioning, billing, terms of service, onboarding enrolment, org mapping, invoicing — is completed without creating a poor customer experience through administrative gaps.

// FREQUENTLY ASKED QUESTIONS