Nine Starting Positions: Find Yourself on the Grid Before You Write a Single Line of Scope

*Series: Building the AI-native enterprise — a practitioner's playbook for scaling Cloud, AI and Agentic AI practices*

*Post 2 of 10 | aiorbitx.com*

*Industries covered across this series: BFSI, healthcare, pharma, manufacturing, retail, logistics.*

01 — Executive hook

In Post 1, I introduced a rule most engagements break: score customer maturity and firm maturity separately, because they're rarely equal. This post makes that rule concrete.

There are exactly nine recognizable starting positions two organizations can be in relative to each other — and I can tell you, within the first thirty minutes of a discovery call, which one I'm looking at. So can you, once you know what to look for.

The nine positions aren't theoretical. They're the nine combinations you get when you cross three types of firm against three types of customer. Each one has a characteristic failure mode, a characteristic success pattern, and a specific engagement model that fits it — and a different one that will quietly wreck it.

02 — The problem

Nobody starts a program by naming their starting position out loud.

Instead, both sides default to their own comfort zone: the firm pitches the engagement model that worked for its last client, the customer scopes the RFP the way their last vendor taught them to. Neither step maps to what's actually true about this pairing.

A firm that's genuinely AI-anchored, walking into a greenfield healthcare client, will over-engineer a governance layer the client isn't ready to staff. A greenfield firm chasing an AI-anchored BFSI client will under-deliver against expectations the client didn't realize were unusually high. Neither failure is about technology. Both are about skipping the fifteen minutes it takes to name the starting position.

The pattern repeats across every industry I work in. A pharma company with a mature GxP-validated data platform hires a partner still running ad hoc AI pilots — and spends the first six months teaching the partner its own compliance requirements. A manufacturer with no cloud practice at all hires a firm that only knows how to extend existing cloud foundations — and the engagement stalls at the infrastructure layer before a single model is trained. The industry changes. The mismatch pattern doesn't.

03 — The maturity question

Post 1 asked you to score four maturity levels across three practice areas. This post asks a simpler, faster question you can answer in one sitting: which of three types are you, on each side of the relationship?

• Greenfield — no formal cloud, AI, or agentic practice yet. Pilots, if any, are one-off. No funded CoE. No repeatable delivery pattern. Governance exists on a whiteboard, not in a cadence.
• Cloud-anchored — a mature, funded cloud practice exists. AI and agentic AI are still ad hoc, bolted on project by project. The cloud CoE is real; the AI practice is aspirational.
• AI-anchored — mature cloud and AI/ML practices exist. Agentic AI specifically is still the newest, least-formed layer — but the foundation is solid and the governance muscle is already built.

Score yourself as one of these three. Then score whoever you're partnering with, independently. Cross the two, and you land on one of nine positions.

One honest rule before you score: don't round up. A cloud practice that runs on heroics and tribal knowledge is not Cloud-anchored — it's Greenfield with better infrastructure. An AI practice that has shipped two pilots and a dashboard is not AI-anchored — it's Cloud-anchored with ambition. The framework only works if both sides name what's actually true.

04 — The framework

The nine positions

Crossing three customer types against three firm types gives nine combinations. Each has a name, a characteristic dynamic, and a default risk.

Nine positions, but they collapse into four practical patterns:

• Aligned-Greenfield (Position 1): both sides build the relationship and the practice at the same time. Slowest start, lowest risk of mismatch. The danger here isn't failure — it's drift. Without a shared governance anchor, both sides improvise in parallel and diverge quietly.
• Firm-led (Positions 2, 4, 7): the firm mentors a less mature customer. Fastest path to delivery, highest risk of over-selling scope the customer can't sustain. The firm's job is to raise the customer's capability, not just deliver against it.
• Customer-led (Positions 3, 6): the customer brings the firm up to speed. Requires unusual honesty from the firm about what it doesn't yet know. The failure mode here is the firm performing competence it doesn't have — and the customer only discovering this six months in.
• Aligned-Advanced (Positions 5, 9): both sides are credible peers. The work shifts from capability-building to co-design. The risk is complacency — two mature organizations assuming alignment without checking it.

What each position actually looks like in the room

Position 1 — Both greenfield

The discovery call feels collaborative and open. Neither side has strong opinions about the delivery model because neither has built one. The risk is that this openness gets mistaken for alignment — it's actually the absence of constraint. First move: agree on a governance charter before you agree on a delivery timeline.

Position 2 — Firm behind

The customer's technical team asks questions the firm's delivery lead can't answer fluently. The firm compensates by talking about roadmap and vision. The customer's CTO starts routing questions to their own internal team instead of the partner. First move: the firm needs to name the gap honestly, or the customer will name it for them — usually in a steering committee.

Position 3 — Firm far behind

Rare, but it happens — usually when a large enterprise with a mature internal AI CoE hires a boutique firm for a specific capability, not for practice leadership. The engagement works only if the firm's role is explicitly scoped as execution, not advisory. The moment the firm tries to lead on architecture or governance, the relationship breaks.

Position 4 — Firm-led build

The most comfortable position for most IT firms. The customer is ready to be led. The risk is that "ready to be led" becomes "dependent on being led" — and the firm builds a practice the customer can't sustain without them. First move: design for handover from day one, not as an afterthought at month eighteen.

Position 5 — Integrate together

The most common pairing in the market right now — a cloud-mature customer and a cloud-mature firm, both trying to figure out AI together. The dynamic is genuinely collaborative, but it can stall when neither side wants to admit they're learning in real time. First move: name the learning explicitly. Build a joint experimentation cadence, not a delivery cadence.

Position 6 — Customer ahead

The customer's AI team is more advanced than the firm's. This is more common in BFSI and pharma than most firms admit — regulated industries have been building internal AI governance for years. The firm's value here is delivery capacity and cross-industry pattern transfer, not AI expertise. First move: the firm should stop leading on AI architecture and start leading on delivery execution.

Position 7 — Firm-led, watch pace

The firm is AI-anchored; the customer is greenfield. The risk is that the firm's default delivery patterns — designed for customers who already have cloud foundations — are too advanced for where the customer actually is. First move: strip the engagement back to the minimum viable governance layer before adding AI capability on top.

Position 8 — Firm mentors AI layer

A solid cloud foundation on both sides, with the firm ahead on AI. This is a high-trust position — the customer has already proven it can build and sustain a practice, so the firm's job is to accelerate, not to build from scratch. First move: audit what the customer's cloud CoE already does well and extend it, rather than replacing it with the firm's own patterns.

Position 9 — Both advanced, agentic is the gap

Both sides have mature cloud and AI practices. The real frontier is agentic AI — and neither side has fully solved it yet. This is the most intellectually honest position in the framework: two advanced organizations admitting they're both learning. First move: establish a joint agentic AI lab with shared IP terms before the first sprint.

05 — Operating model *(previewed here, full treatment in Post 6)*

Your position determines which operating model is even viable. A Greenfield-Greenfield pairing (Position 1) cannot sustain a federated model yet — there's no CoE on either side to federate around. A Position 9 pairing can go straight to federated, and probably should. Positions 4 and 7 typically suit a hub model where the firm's CoE acts as the central capability anchor. Post 6 maps all nine positions to a recommended operating model with the actual team-relationship diagrams.

06 — Organization *(previewed here, full treatment in Posts 6 and 9)*

Who staffs the relationship changes by position. Firm-led positions (2, 4, 7) need the firm to bring a client-facing enablement lead — not just delivery engineers. Customer-led positions (3, 6) need the customer to designate a technical counterpart who is empowered to teach, not just approve. Position 5 and 9 pairings need a joint governance forum with real decision rights on both sides, not a steering committee that only reviews status. Post 9 covers the full RACI by position.

07 — Delivery *(previewed here, full treatment in Post 7)*

Firm-led positions usually suit a managed-practice or staff-augmented delivery model — the firm brings the pattern, the customer provides the domain context. Customer-led positions often work better as a co-innovation lab, where the customer's internal team leads on architecture and the firm provides delivery capacity and cross-industry reference patterns. Position 5 — the most common pairing — typically starts as staff-augmented and evolves toward a joint CoE model as AI capability matures on both sides. Post 7 walks through the fit by position with realistic transition timelines.

08 — Technology *(previewed here, integrated into Posts 4, 5 and 7)*

A Greenfield pairing needs one well-governed reference architecture — not a platform, not a catalog, just a single agreed pattern that both sides can execute against. An AI-anchored pairing (Position 9) is usually ready for a self-service accelerator catalog on day one. Tooling maturity should track the position, not the vendor's default stack. And it stays sensitive to sector rules: HIPAA and GxP evidence trails in healthcare and pharma, PCI DSS and RBI/SOX-class controls in BFSI, OT/IT boundary governance in manufacturing.

09 — Economics *(previewed here, full treatment in Post 8)*

Firm-led positions typically carry a heavier upfront enablement cost — the firm is building capability the customer doesn't yet have, and that cost has to land somewhere. In practice it either gets absorbed into the firm's margin, amortized across a multi-year engagement, or explicitly scoped as a capability-building workstream with its own budget line. Customer-led positions shift more of that cost — and the associated leverage — to the customer. Position 5 pairings often have the most negotiable economics, because both sides are contributing real capability. Post 8 breaks down realistic budget bands by position and industry.

10 — Business *(previewed here, full treatment in Post 8)*

Value narratives differ by position. Firm-led engagements sell on de-risked delivery — the customer is buying certainty that the firm has done this before and won't learn on their budget. Customer-led engagements sell on speed-to-scale — the customer is buying delivery capacity and cross-industry pattern transfer, not expertise they already have. Position 5 and 9 pairings sell on co-innovation — the value is in what neither side could build alone.

11 — Roadmap

• Days 1–30: use the 3×3 grid above to name your position honestly, on paper, with your counterpart present. Both sides score independently first, then compare. Disagreement on the position is itself diagnostic — it usually means one side is rounding up.
• Days 31–60: read the specific post this position points you to (Post 3 for Greenfield, Post 4 for Cloud-anchored, Post 5 for AI-anchored) before the next scoping conversation. Firm-led and customer-led positions often mean each side reads a different post — that's expected, not a mistake.
• Days 61–90: revisit the position at the end of the first delivery cycle. Positions shift — if they aren't shifting toward alignment, the operating model isn't working. A firm-led position that hasn't moved toward Position 5 after ninety days of delivery is a signal, not a coincidence.

12 — Executive checklist: what a CIO/CTO should do Monday morning

☐ Name your organization's type — Greenfield, Cloud-anchored, or AI-anchored — without rounding up

☐ Ask your delivery partner to do the same, out loud, in the same meeting

☐ Identify your combined position from the 3×3 grid above

☐ Confirm the operating and delivery model on the table actually fits that position

☐ Flag any position where the firm is Greenfield and the ask is platform-scale — the highest-risk pattern in this framework

☐ If you're in a customer-led position (3 or 6), confirm the firm has explicitly acknowledged it — not just nodded along

☐ If you're in Position 9, confirm you have joint IP terms agreed before the first agentic AI sprint

13 — Quick check

One-question self-check: when you described your AI practice to your last delivery partner, did they push back on your self-assessment — or did they agree with everything you said?

If they agreed with everything, one of two things is true: you assessed yourself accurately, or the partner was telling you what you wanted to hear. The second is more common. A partner who never challenges your maturity self-assessment is almost certainly in a firm-led position and doesn't want to risk the engagement by naming it.

The full multi-dimension self-assessment, with automatic scoring and the auto-plotted maturity grid, is in the Post 1 companion workbook — Practice Maturity Self-Assessment.xlsx.

14 — Closing thought

Maturity level (Post 1) tells you how formal a practice is. Starting position (this post) tells you what kind of practice it is and which direction it needs to grow. The two are related but not the same — a Cloud-anchored organization can be L1 or L4 depending on how disciplined that cloud practice actually is.

Use both lenses together, and you'll walk into a scoping conversation already knowing more about the fit than most RFPs are designed to surface. The nine positions aren't a scoring exercise — they're a shared language. Once both sides are using it, the conversation changes from "here's what we propose" to "here's where we actually are, and here's what that means for how we work together."

That's a different conversation. It's a better one.

*Next in the series — Post 3: Starting From Zero — the tactical playbook for a Greenfield practice: first hire, first pilot, first governance artifact.*