Partnerships Architect

Partnerships Architect

End-to-end strategic partnership design and scaling: partnership type selection (tech integration vs channel vs OEM vs strategic), deal structures, partner program design (tiers, benefits, requirements), partner evaluation, and ROI modeling that justifies (or kills) a partnership investment.

This skill is provider-agnostic and works across SaaS, infrastructure, marketplace, and platform companies.

When to use this skill

Partnership types — the decision tree

Five primary partnership types. Different goals; different structures; different success metrics.

Copy code
What's the primary goal of this partnership?

Grow our distribution reach
├── Customer-pays-them, they-pay-us → Reseller / Distributor / VAR
├── Customer-pays-us, we-pay-them → Affiliate / Referral
└── Joint sale to mutual customer → Co-sell

Embed our product in their offering
├── Customer doesn't see us (white-label) → OEM
├── Customer sees us as embedded → Embedded ISV / Powered-by
└── We're an option in their marketplace → Marketplace listing

Combine our product with theirs (better together)
├── Pre-integrated, certified → Tech / Integration Partner
├── Bundled offering → Solution Partner
└── Joint product (rare) → Joint Venture

Build market presence together
├── Joint events, content, PR → Co-marketing Partner
├── Industry positioning → Strategic Alliance
└── Standards / consortium → Standards Partner

Achieve a specific strategic goal
├── Block a competitor → Defensive partnership
├── Enter a new market → Market entry partnership
└── Acquire capability → Strategic alliance (often pre-acquisition)

See references/partnership-types.md for each type in depth: economic structure, contract patterns, KPIs, when each works / fails.

Partner evaluation framework

Not every potential partner is worth the investment. Use this framework before committing.

Six evaluation dimensions

Scoring rubric

• 5 — strong yes
• 4 — yes with minor caveats
• 3 — mixed; substantial uncertainty
• 2 — weak; significant concerns
• 1 — no; deal-breaker

Total 25-30: green-light; invest with confidence Total 18-24: yellow; structure carefully; small pilot first Total < 18: red; decline or substantially restructure

Use scripts/partner_evaluation_scorer.py --partner partner.yaml to score a specific potential partner.

Killer questions to ask

Before signing any significant partnership:

1. What does success look like in 12 months? If both sides can't articulate the same answer, you don't have alignment.
2. What's their commitment level? Headcount assigned? Budget? Executive sponsorship?
3. What's the realistic pipeline in next 12 months? Specific accounts? Or vague "we have customers"?
4. Who's the day-to-day owner on each side? Names + tenure + reporting line.
5. What happens if we don't hit our shared metrics? Course-correct? Wind down? Renegotiate?

If you can't get clear answers, the partnership is wishful thinking.

Partnership deal structures

Different deal types call for different structures. Standard patterns:

Structure A: Standard reseller agreement

• Term: 1-3 years, auto-renew
• Discount: per published tier matrix
• Exclusivity: usually non-exclusive
• Termination: 90-day notice both sides
• Use when: typical channel relationship

Structure B: Co-sell agreement (mutual customer)

• Term: 1-2 years
• Compensation: shared commission OR referral fee
• Joint marketing commitment: optional
• Use when: complementary offerings; existing or target shared customers

Structure C: OEM agreement

• Term: 3-7 years (long; relationship-heavy)
• Royalty / rev-share: % of partner revenue OR per-instance fee
• Exclusivity: often partial (specific use case / market)
• Source code escrow: usually required
• Termination: complex (typically 12-24 months notice; transition rights)
• Use when: deeply embedded technical relationship; high mutual investment

Structure D: Strategic alliance

• Term: open-ended; reviewed annually
• Resources committed: explicit (e.g., 2 FTEs per side, $X budget per year, joint roadmap session quarterly)
• Governance: steering committee (executive sponsors meet quarterly)
• Specific deliverables: joint product features, joint customer wins, joint thought leadership
• Use when: 5+ year strategic relationship; not transactional

Structure E: Tech / integration partnership

• Term: 1-3 years
• Compensation: typically none direct; mutual value from joint customers
• Certification process: defined (testing, documentation)
• Marketplace listing: typically included
• Co-marketing: optional but common
• Use when: integration creates joint customer value; no direct revenue flow

See references/partnership-deal-structures.md for the full deal-structure templates with negotiation guides.

Partner program design

When you scale beyond a few partners, you need a program.

Three pillars of a partner program

Standard program elements

• Partner agreement (master): tenant-of-the-relationship
• Tier structure (Authorized → Silver → Gold → Platinum): different benefits + requirements per tier
• Deal registration: protect partner-developed opportunities
• Certification program: train + test partners on your product
• Partner portal: deal reg, MDF, training, marketing materials, lead sharing
• MDF (Marketing Development Funds): co-funded marketing
• Channel manager(s): 1 per 10-15 active partners
• Annual partner conference: community building + recognition

See references/partner-program-design.md for the full program template including tier definitions, benefit / requirement matrices, and the "Year 1 / Year 2 / Year 3" maturity model.

Use scripts/partner_program_designer.py --org-spec org.yaml to generate a baseline program design based on company stage + ICP + target partner volume.

Partnership ROI modeling

Partnerships consume real investment. Headcount, MDF, technology, executive time. Model the ROI before committing.

ROI model template

Copy code
3-year cumulative partnership P&L

Year 1 (investment year):
  Revenue from partnership: $X
  Costs:
    Partnership manager headcount: $200k
    Engineering integration (one-time): $300k
    Marketing / co-launch: $50k
    Partner enablement (content, training): $50k
    Travel / events: $30k
    TOTAL Y1 cost: $630k
  Y1 net: $X - $630k

Year 2:
  Revenue from partnership: $Y (growth)
  Costs:
    Partnership manager: $200k
    Engineering ongoing: $100k
    Marketing: $80k
    Enablement: $30k
    Travel / events: $40k
    TOTAL Y2 cost: $450k
  Y2 net: $Y - $450k

Year 3:
  Revenue: $Z (mature)
  Costs: $400k (stable)
  Y3 net: $Z - $400k

3-year cumulative net: ($X + $Y + $Z) - $1,480k

If 3-year cumulative net is negative, the partnership doesn't pay back. Common reasons:

• Revenue estimates too optimistic
• Forgotten costs (executive time, opportunity cost)
• Partner under-performs on commitments
• Market shift makes the partnership less relevant

Use scripts/partnership_roi_modeler.py --partnership partnership.yaml for the full model.

Clarify First

Before designing the partnership, confirm these inputs. If any is unknown or vague, ASK — do not assume:

• Partnership goal + type — distribution, embed (OEM/ISV), better-together (tech/integration), market presence, or strategic (selects the decision-tree branch and deal structure)
• Single deal vs scaled program — structuring one partnership vs designing tiers/portfolio (determines deal-structure vs program-design output)
• Partner specifics for evaluation — their pipeline expectation, commitment, and credibility (feeds the 6-dimension scorer and green/yellow/red call)
• ROI inputs — expected revenue and the real costs (headcount, integration, MDF) (drives the 3-year P&L payback)

Stop rule: ask only the 2-3 that most change the output. If the user says "just draft it," proceed and list your assumptions at the top of the deliverable.

End-to-end workflows

Workflow: Evaluate a new partner opportunity

1. Initial conversation with potential partner — understand their pitch
2. Gather data for evaluation: their company, their pipeline expectations, their commitment
3. Score using evaluation framework — scripts/partner_evaluation_scorer.py
4. If score < 18: decline or pilot-scope only
5. If score 18-24: pilot scope (3-6 months, limited investment)
6. If score 25+: standard partnership agreement; full investment

Workflow: Stand up a partner program

1. Define target partner profile (TPP): ideal partner attributes (size, vertical, region, capability)
2. Pick partnership types in scope (resellers? OEM? tech partners?)
3. Design tier structure — scripts/partner_program_designer.py
4. Build foundational tools: partner portal, deal-reg, certification
5. Recruit pilot cohort (3-5 partners) — manual, high-touch
6. Iterate based on pilot feedback — usually 6 months
7. Scale recruitment — content marketing, outbound, partner events
8. Add channel managers as portfolio grows (1 per 10-15 active partners)

Workflow: Structure a specific partnership deal

1. Identify partnership type using the decision tree
2. Score the partner — scripts/partner_evaluation_scorer.py
3. Pick deal structure matching the type
4. Model ROI — scripts/partnership_roi_modeler.py
5. Draft term sheet (key business terms)
6. Negotiate — alignment on commitment, timelines, exit
7. Legal review + contract — business-growth/contract-and-proposal-writer
8. Internal approval — Deal Desk + CRO/CFO/CEO depending on scale

Workflow: Audit existing partner portfolio

1. List all active partners + key data: revenue, costs, deals, tier
2. Score each against evaluation framework — scripts/partner_evaluation_scorer.py
3. Identify low-ROI partners — bottom quartile by contribution per investment hour
4. Decide per partner:
   • Invest more (top quartile, expand commitment)
   • Maintain (middle, status quo)
   • Wind down (bottom; explicit timeline to exit gracefully)
5. Quarterly review of portfolio with CRO

Anti-patterns

• "Strategic" partnership that's actually transactional. If the only thing exchanging is money, it's transactional; call it that. Strategic means joint goals + shared roadmap + executive commitment.
• Partner stack with no portfolio strategy. Signing every partner that asks. Quality > quantity. 10 productive partners beat 100 zombies.
• No partnership owner. Partnership exists in nobody's job. It withers.
• Resource asymmetry. You commit 3 FTEs; they commit 0.5. The partnership skews to their convenience.
• Promises without commitments. "We'll do joint webinars." When? With what budget? Whose role to organize?
• Open-ended exclusivity. "Exclusive in this region forever" without performance gates. Lose flexibility for nothing in return.
• OEM deal with no source-code escrow. Customer-impact risk if your company goes away.
• Strategic alliance with no governance. No quarterly review = no executive engagement = partnership drifts.
• Partner program with no enablement. Partners can't sell what they don't understand.
• Tier benefits that aren't worth tier requirements. Partners don't advance because there's no incentive.
• Co-marketing dollars wasted on activities without pipeline. Great event, zero attribution.

Tooling outputs

All scripts: stdlib only, argparse CLI, JSON or markdown output.

References

• partnership-types.md — 5 partnership types in depth + economic structure + when each works
• partnership-deal-structures.md — deal templates per type + negotiation guides
• partner-program-design.md — full program design with tier matrices + maturity model

Related skills

• business-growth/channel-economics — per-deal financial mechanics underneath partnership structure
• business-growth/deal-desk — per-deal approval mechanics for partner-mediated deals
• business-growth/pricing-strategy — pricing flexibility / floor for partner deals
• business-growth/contract-and-proposal-writer — partner contracts (MSA, partner agreement, OEM agreement)
• c-level-advisor/cs-cro-advisor — strategic-level partnership decisions
• c-level-advisor/cs-ceo-advisor — board-level alliance decisions