Roche on Strategy
Strategy for consequential decisions and changing enterprises.
Essay
No. ROS 0002
Published · 14 JUL 2026

14 JUL 2026
§ ROS 0002

The Ecosystem

A partner network becomes an ecosystem only when shared value, authority, incentives, and execution are intentionally designed.

Last revised · 30 JUL 2026

Partner motions do not fail at the partner. They fail at the seam between product, sales, and finance. A framework for aligning the operating model to the ecosystem strategy - before the ecosystem strategy is asked to carry weight it cannot bear.

The wrong diagnosis

When a partner ecosystem underperforms, the postmortem almost always lands in the same place: the partner team needs more headcount, better tooling, a fresher tier structure, a co-sell motion, a certification track. The remedy gets scoped to the function that carries the number. The number does not move. Twelve months later, a new head of partnerships is hired to run the same play with slightly different vocabulary.

The diagnosis is wrong. Partner motions rarely fail because the partner team is under-resourced or under-skilled. They fail because the operating model around them - how product decides what to build, how sales decides what to sell, how finance decides what to count - was never adjusted to a world in which revenue arrives through someone else's motion. The ecosystem strategy is asked to carry weight the operating model refuses to bear. That is not a partnerships problem. That is an executive problem.

What an operating model actually is

An operating model is not an org chart. It is the set of standing decisions a company has made - often silently - about how work gets prioritized, how money moves, how accountability is assigned, and how conflicts get resolved when two functions want incompatible things. Most companies can articulate their strategy on a slide. Very few can articulate their operating model without describing the org chart, which is the giveaway that the operating model has not actually been designed.

A partner ecosystem is a distribution and value-creation choice that reaches into every one of those standing decisions. It changes what product should build, because the integration surface becomes as strategic as the feature surface. It changes what sales should sell, because a deal sourced by a partner has different economics and different risk than a direct deal. It changes what finance should count, because attribution, margin, and forecast confidence all move. If those standing decisions do not change, the ecosystem strategy has been declared but not adopted.

The four seams where partner motions fail

In practice, the failure shows up at four specific seams. None of them are the partner relationship itself. All of them are internal.

Product ↔ Partner. The product roadmap is set against the direct customer's request list. Partner-driven requirements - integrations, embedded experiences, white-label surfaces, API contracts stable enough to build a business on - arrive as tickets and are prioritized against features. They lose, quietly, every quarter. The partner team learns not to promise anything on the roadmap. The partner learns not to build.

Sales ↔ Partner. Compensation is the tell. If a rep is paid the same on a partner-sourced deal and a self-sourced deal, the operating model is neutral to partners. If a rep is paid less on partner-sourced deals - or, more common, the same number but with more perceived friction - the operating model is hostile to partners. The partner team writes a rules-of-engagement document; the field ignores it; the CRO intervenes case by case. That is not a policy problem. It is an operating model that has not decided what a partner-sourced dollar is worth.

Finance ↔ Partner. Finance runs on attribution and margin. Ecosystem revenue often arrives with murky attribution (was it sourced, influenced, or fulfilled by the partner?) and different margin structure (referral fees, rev-share, marketplace take rates, co-sell splits). If finance has no category for partner revenue distinct from direct, the forecast punishes it and the board reads the wrong story. The ecosystem strategy then gets defunded not because it failed, but because it was measured as if it were direct sales that underperformed.

Executive ↔ Partner. The final seam is the one that determines the other three. If the ecosystem strategy lives inside the partnerships function, it has the authority of one VP. If it lives inside the operating model, it has the authority of the CEO. Every durable ecosystem I have watched work at scale has an executive who treated the ecosystem as a first-class distribution channel and rewired the operating model accordingly. Every ecosystem that stalled had a VP of Partnerships trying to negotiate that rewiring from below.

A framework: the Ecosystem Operating Model

The Ecosystem Operating Model is a specific set of decisions the executive team makes - once, explicitly, in writing - about how the ecosystem strategy is carried by the rest of the company. It has four components, mapped one-to-one against the seams above.

1. Product commitment. A defined share of the roadmap, quarter by quarter, is reserved for the ecosystem surface: integration stability, partner-facing APIs, embedded experiences, developer tooling. The share is set at the executive level and defended against direct-customer escalations by the same governance that defends any strategic bet. "As much as we can spare" is not a commitment; it is a decline dressed up.

2. Commercial architecture. Compensation, discount authority, and deal desk rules are rewritten so that a partner-sourced dollar and a direct dollar are worth the same to the rep who carries the number. If ecosystem revenue is genuinely more valuable (higher retention, larger expansion, lower CAC), the rep is paid more on it. The commercial architecture is the operating model's most legible expression of what the strategy actually values.

3. Financial categorization. Finance recognizes partner revenue as a distinct category with its own attribution logic, margin profile, and forecast discipline. The board sees three lines, not one: direct, partner-sourced, partner-fulfilled. The ecosystem strategy is then measured against the plan that was set for it, not against the plan that was set for direct sales.

4. Executive ownership. One executive - typically the CEO or COO, occasionally the CRO if the CRO's mandate genuinely spans channel - owns the ecosystem outcome. The head of partnerships runs the function; the executive owns the cross-functional decisions that determine whether the function can succeed. This is not delegation with reporting lines. It is accountability that cannot be handed down.

The executive question this answers

The reason ecosystem strategy so often stalls at the executive table is that it gets presented as a partnerships question - tier structures, partner personas, program mechanics - when the executive team is being asked to answer an operating model question: are we willing to change how product, sales, and finance make decisions in order to make ecosystem revenue real? A partnerships deck cannot answer that question because it does not ask it.

The Ecosystem Operating Model reframes the conversation. It sets aside the tactical program and puts the four commitments on the table, each with its cost, each with its consequence for the functions being asked to change. That is the conversation an executive team can actually have, and the conversation the partner team cannot force from below.

Where to start on Monday

If you run an ecosystem function, the useful first move is not another program redesign. It is a one-page diagnostic against the four seams: which are aligned, which are neutral, which are actively hostile. Bring that page to your executive sponsor, not your peers. The seams that are hostile will not become neutral through partner team effort; they will become neutral through executive decisions that only the executive can make.

If you run the company, the useful first move is to stop treating the ecosystem as a function to be managed and start treating it as an operating model choice to be made. The strategy is the easy part. The operating model is where the strategy either survives contact with product, sales, and finance - or does not. Everything downstream of that decision is execution. Everything upstream of it is a slide.