Domain 2: Strategy & Positioning

TL;DR. Where the company decides what to be in the market — the most agent-resistant domain. Anchor framework: April Dunford's 5-component canvas (300+ B2B tech companies advised). Tools that win: Writer.com (regulated industries) or Custom Claude Project (lean teams) for brand voice; Wynter ($799+/mo) for B2B message validation; Klue/Crayon for competitive intel. Canonical case: Brand.ai — enterprise brand-compliance costs $5M/yr → small fraction; one copywriter manages 600 pieces of content. What changed in v3: added Dunford March 2026 Lenny's Newsletter quotes, 5 named cases (Userlist, Influ2, Workday, Superhuman, Brand.ai), 8-agent Brand Governance Agent build playbook (Animalz pattern), Wynter-vs-SparkToro-vs-PyMC-Labs sequencing, and Klue-vs-Crayon-vs-Custom-Clay comparison.

"I believe disagreement about what to position against is the root of most weak positioning." — April Dunford, A guide to advanced B2B positioning (Lenny's Newsletter), Mar 10, 2026

"in an effective strategic narrative, your buyer's enemy is an old mindset that has become a road to ruin." — Andy Raskin, LinkedIn, Oct 2023

See also: Domain 0 (AgentOps) for the Brand Governance Agent that operationalizes this domain's brand voice spec, Domain 3 (Content) for execution at velocity, Domain 5 (AEO/GEO) for positioning that survives LLM-citation contexts, Domain 7 (Customer Intelligence) for synthetic + live message testing, Domain 1 (Sensing) for sales-call mining → positioning research.

Definition and Scope

Where the company decides what to be in the market. ICP definition, segmentation, messaging architecture, value-proposition design, competitive counter-positioning, brand voice codification, pricing narrative.

This domain is the most agent-resistant of the eight. Agents are extraordinary research and synthesis partners, but the actual call about what to be, what category to live in, what to attack, what to abandon, is irreducibly human work. It requires taste, judgment, and the willingness to be wrong publicly.

Why It Matters Now

April Dunford, the world's leading authority on B2B tech positioning, has worked with over 300 B2B tech companies and has consistently found that the root of weak positioning is disagreement among the founder, sales, marketing, and product about who they actually compete with. In an agentic world, this gets worse, not better, because agents will execute brilliantly against bad positioning, producing 10x the volume of confused messaging.

The structural shift: with execution velocity unlocked, positioning becomes the bottleneck. Bad positioning used to manifest slowly (over a quarter or two of disappointing pipeline). Now it manifests in two weeks of spam-flagged sender reputation and dropped meetings.

Sub-Domains

2.1 ICP Definition & Segmentation

• Tier 1 / Tier 2 / Tier 3 account stratification
• Best-fit-customer characterization (not biggest-spender)
• Persona-within-account decomposition (champion, economic buyer, technical evaluator, end user)
• Negative ICP, who you're explicitly NOT for

2.2 Positioning Architecture

• Competitive alternatives (what does the prospect compare you to?)
• Unique attributes (what only you have?)
• Value (why does that matter to a buyer?)
• Best-fit customer (who cares most about this value?)
• Market category (what do you call yourself?)

2.3 Messaging Architecture

• Category narrative
• Per-persona value props
• Per-vertical adaptations
• Proof point inventory (case studies, stats, social proof)
• Objection handling library

2.4 Competitive Counter-Positioning

• Per-competitor battlecards
• Win/loss pattern analysis
• "Reasons to switch" narratives
• Category-defining moves (when to attack, when to ignore)

2.5 Brand Voice Codification

• Tonal guidelines (cinematic vs. clinical vs. conversational)
• Vocabulary boundaries (terms you own, terms you avoid)
• Sentence structure patterns
• Visual identity rules (color, type, layout)
• The "Brand Governance Agent" knowledge base

2.6 Pricing Narrative

• Anchor points
• Packaging logic
• Discount discipline
• Value-justification language

Best Practices in 2026

Run positioning as a cross-functional exercise, not a marketing one. Dunford's premise: positioning cannot live in the marketing department alone. If the CEO, sales lead, and product lead aren't in the room providing input, marketing is left guessing about what makes the product special and who it's actually for. This is more true with agents in the loop, not less.

Define the competitive set carefully. Each function in a B2B company has its own bias about who they compete with. Marketing watches competitors with big media spend. Sales watches who shows up on shortlists. Product watches who's shipping similar features. Founders watch who's raising money. They're often looking at four different sets. Reconcile this before you do anything else.

Codify the brand voice into an agent-readable specification. This is the single highest-leverage strategy move in an agentic stack. The "Brand Governance Agent", fine-tuned on your existing high-performing content, ontological dictionary, and tonal guidelines, becomes the gatekeeper that prevents the execution plane from shipping garbage. Without it, you'll produce 100x the volume of off-brand content.

Run continuous competitive intelligence, not periodic refreshes. Agentic systems can monitor competitor pricing pages, hiring patterns, and messaging drift weekly without human effort. Use it. Quarterly competitive refreshes are a relic of pre-agentic ops.

Treat pricing as positioning, not a finance exercise. Pricing communicates positioning more strongly than any tagline. A 5x price differential signals what category you're in.

Tools & Platforms

Positioning & Strategy Frameworks (Frameworks > Tools)

• April Dunford's positioning canvas (5 components)
• Andy Raskin's "Strategic Narrative" framework
• Patrick Campbell (ProfitWell/Paddle) on pricing
• Wynter, message testing with B2B audiences ($299–$1,000+/mo)

Customer Research Platforms

• UserInterviews.com, recruiting research participants
• Wynter, message testing, 100+ verified B2B audience pools
• SparkToro, audience research, "what does your audience read/watch/listen to?"
• PyMC Labs, synthetic consumer panels for B2B (see Domain 7)

Brand Voice & Governance

• Writer.com. Enterprise brand governance, style guide enforcement, terminology management. Strongest for regulated industries.
• Jasper IQ. Brand voice training from existing content samples; multi-brand support.
• Claude Projects. Lightweight brand voice via project-level system prompts and knowledge bases.
• Custom-built Governance Agent, fine-tuned LLM on your style guide, BEKG/ontology, high-performing content samples.

Competitive Intelligence

• Klue. Enterprise competitive intelligence platform
• Crayon. Competitive intelligence + battlecards
• Kompyte. Lower-cost competitive monitoring
• Custom Clay/n8n workflows. Often outperforms commercial tools for specific competitor lists

Notable Practitioners & Frameworks

• April Dunford; Obviously Awesome, Sales Pitch, 5-component positioning canvas. 300+ B2B tech companies advised over 10 years. Recent: "A guide to advanced B2B positioning". Lenny's Newsletter, March 2026; "Why April Dunford Separates B2B Positioning from Product Strategy". Dreamdata, 2025.
• Andy Raskin. Strategic Narrative framework. Don't attack competitors; attack an old mindset and frame competitors as relics of it. Clients: Salesforce, Gong, Dropbox, Square, Uber, IBM, Turing. "The power of strategic narrative". Lenny's Newsletter. Real-time case: Influ2 narrative overhaul, March 2025.
• Christopher Lochhead; Play Bigger; departed Play Bigger to focus on Category Pirates. "How to become a category pirate". Lenny's Podcast.
• Patrick Campbell. Pricing strategy; ProfitWell Report archive (Paddle). Hard numbers from ~30K-subscriber experiments: B2B case studies lift WTP 10–15%; positive brand perception lifts WTP 16–41% above median; 5+ integrations lift WTP 20%+.
• Bob Moesta. Jobs-to-be-Done. "The ultimate guide to JTBD". Lenny's Podcast; new podcast The Circuit Breaker.
• Anthony Pierri (Fletch PMM, not "Halfwit"); 400+ B2B SaaS homepages advised. Frame: choose between "use case" and "category" homepage strategies; never both. "How to Master B2B Positioning". Exit Five #323.

Named Case Studies

Tools & Platforms: Updated 2026

Brand Voice: Writer.com vs. Jasper IQ vs. Custom Claude Project

Decision rule: Compliance-driven → Writer. Multi-brand toggling → Jasper. 8 specialized governance agents inspecting one document with full prompt control → Claude Projects + Skills.

B2B Message Testing (sequence them)

Sequence: SparkToro for discovery → PyMC Labs for cheap pre-screening → Wynter for final validation before launch.

Competitive Intel: Klue vs. Crayon vs. Custom Clay/n8n

Tactical Playbooks

Playbook A. Build a Brand Governance Agent in Claude Projects (8-agent parallel)

System prompt skeleton:

You are a [voice/grammar/terminology/legal/etc.] auditor for [Brand X].
Read the attached style guide (knowledge file).
For each violation, output: line number | violation type | severity (1-3) | quoted phrase | suggested replacement.
Score the document 0–100 on your single dimension.
Do NOT flag dimensions outside your assigned scope.

Style-guide knowledge structure (uploaded to Project):

• Voice axes (e.g., "cinematic vs clinical 7/10")
• Vocabulary boundaries (terms-we-own / terms-we-avoid lists)
• Sentence structure patterns (banned: "In today's fast-paced world…")
• Persona table (champion, EB, technical evaluator)
• Failure-mode examples (5–10 annotated bad outputs)

Run pattern: orchestrator agent dispatches to 8 specialists in parallel, aggregates scores into a single Brand Governance Score. Threshold: <85 = block publish; 85–94 = human review; ≥95 = auto-publish.

Reference impl: anthropics/skills brand-guidelines + the Animalz Claude-Code workflow.

Cross-link: Domain 0 AgentOps Playbook A, same architecture viewed from the operations layer.

Playbook B. Dunford's 5-Component Canvas as Fillable Template

Templates: Miro community, Figma community, April's official PDF (gated form on aprildunford.com).

Playbook C. Reconcile Competitive Set Across CEO/Sales/Marketing/Product

90-minute workshop. Each function lists their top 5 competitors independently before the meeting. Compare lists in the room. Dunford's finding: the four lists usually agree on at most 1–2 names. Force-rank the union, then narrow to a single "official" set. Annual cadence; monthly delta review run by a Klue/Crayon/Clay agent.

Cross-References to Mahmoud's Skills

• product-marketing-context, owns the "set up product/audience/positioning context for this project" loop. Domain 2 should link to it as the operationalization of positioning, not restate.
• copywriting, owns page-by-page execution (homepage, pricing, feature pages). Domain 2 stays at the architecture layer.
• competitor-alternatives, owns the listicle/vs/alternative-page execution layer. Cross-link to Domain 5 (these pages drive AEO/GEO listicle citations).
• pricing-strategy, owns Van Westendorp, packaging tiers, monetization. Domain 2 cites Patrick Campbell's WTP numbers and the 5×-as-category-signal rule, then links out.

Industry overlay (Q2 2026)

Key insight: Biopharma positioning is bottlenecked by MLR review, every claim, including AI-generated headlines, gets line-by-line scrutinized for fair balance. Build a pre-MLR governance agent (Veeva Vault PromoMats integration) into the brand layer or you'll burn agency time on rework.

Common Failure Modes

• Marketing-only positioning. The CEO and sales lead don't agree, so the messaging contradicts what's pitched in deals.
• Treating positioning as a deliverable, not a process. A positioning doc that lives in a slide deck and never updates becomes wrong fast.
• Following B2C frameworks for B2B work. Most positioning courses are full of toothpaste and shampoo examples. B2B positioning works differently, multi-stakeholder buying, longer cycles, higher dollar values.
• Mistaking a tagline for positioning. Positioning is a market structure decision. A tagline is a sentence.
• Updating positioning before validating. Synthetic personas + Wynter testing + sales call analysis should validate every material positioning shift before it ships.

KPIs

• Positioning clarity (qualitative: can the team articulate it in one sentence consistently?)
• Sales win rate against named competitors (the proof of positioning)
• Average deal size (positioning failure → pricing pressure)
• Sales cycle length (clear positioning shortens cycles)
• Brand-search volume (over time)
• Inbound-to-outbound ratio (clear positioning generates inbound)

Resources for Deeper Study

YouTube channels & videos

• April Dunford's YouTube channel. Positioning workshops, B2B strategy
• Lenny Rachitsky. Product strategy, includes April Dunford appearances
• Marketing Examined (Alex Garcia). B2C-skewed but excellent positioning teardowns
• Pavilion (channel). Revenue leadership content

Podcasts

• Lenny's Podcast (Lenny Rachitsky), frequent positioning episodes
• In Depth (First Round Capital)
• The Marketing Book Podcast
• Acquired (for category-creation case studies; unmatched depth)

Books

• Obviously Awesome (April Dunford)
• Sales Pitch (April Dunford)
• Play Bigger (Lochhead, Ramadan, Peterson, Maney)
• The Jolt Effect (Matt Dixon)
• Crossing the Chasm (Geoffrey Moore)
• Positioning: The Battle for Your Mind (Ries & Trout), the original

Substacks & Newsletters

• April Dunford's Substack
• Lenny's Newsletter
• Andy Raskin's "The Strategic Narrative"

v3 (shipped Apr 2026)

• Dunford Mar 2026 Lenny's Newsletter quotes verbatim
• Andy Raskin 'old mindset' verbatim
• 5 named cases (Userlist 4× iteration, Influ2 strategic narrative overhaul, Workday $5B archetype, Superhuman 5× pricing-as-positioning, Brand.ai $5M→fraction)
• 8-agent Brand Governance Agent build playbook (Animalz pattern referenced)
• Wynter / SparkToro / PyMC Labs sequencing framework
• Klue vs. Crayon vs. custom Clay comparison
• Industry overlay + cross-references (5 inter-domain + 4 skills)

v4 deferred

• Brand Governance Agent regression-set methodology + violation-rate benchmark
• Pricing-as-positioning case in regulated industries (biopharma 5× rule)

See research-plan.md for the master v3 changelog and v4 forward plan.

Frequently Asked Questions — Domain 2: Strategy & Positioning

Why is positioning the most agent-resistant domain?

Agents are extraordinary research and synthesis partners — they can map competitors, surface positioning gaps, and pressure-test narratives against synthetic personas. But the actual call about what to be in the market requires taste, judgment, and the willingness to be wrong publicly. That call is irreducibly human work. Agents execute against positioning; they don't decide it.

What is April Dunford's 5-component canvas?

Five questions every B2B positioning exercise must answer: (1) competitive alternatives — what would the buyer use if you didn't exist? (2) unique attributes — what only you have? (3) value — why does that matter to a buyer? (4) best-fit customer — who cares most? (5) market category — what frame makes your value obvious? Templates available on Miro and Figma. Dunford has worked with 300+ B2B tech companies using this canvas.

What is a Brand Governance Agent?

A multi-agent system that reviews every public-facing output before publish. The 8-agent parallel architecture (Animalz pattern) runs voice, grammar, banned-terms, legal-claims, persona-fit, sentence-structure, terminology, and factual-grounding auditors simultaneously, then aggregates a 0-100 Brand Governance Score. Threshold: <85 blocks; 85-94 routes to human review; ≥95 auto-publishes. Reference implementation: anthropics/skills brand-guidelines repo.

Should you reposition or remessage?

Almost always: remessage. April Dunford's March 2026 warning: teams reposition far less often than they think. Repositioning changes the market frame; remessaging changes how you describe yourself within the same frame. Disagreement among CEO/sales/marketing/product on the competitive set is the single biggest source of weak positioning — fix that before changing the position itself.