The Most Important Partner

What Greylock Just Admitted About Partnership Value

Greylock Partners raised a $1.5 billion fund this week and made a point of telling everyone that they could have raised a multiple of that figure. The restraint was intentional, and the explanation was specific. Partner Saam Motamedi said the firm's mission is to be "the most important partner to the most important entrepreneurs," and that achieving that requires keeping the number of portfolio companies small. When their partners meet on Monday mornings to review the investment pipeline, the agenda is organized around people's names, not company names. The company often does not exist yet. The bet is on the person.

Greylock is describing a partnership philosophy that most founders claim to believe and almost none actually practice. They believe it in the abstract, the way most people believe in the importance of sleep. They practice something else entirely, and the gap between the philosophy and the practice is where the majority of partnership value goes to die.

The standard founder approach to strategic partnerships runs on the same logic as a sales pipeline: build volume, qualify at every stage, convert the best-fit prospects, and measure success by how many agreements are active at any given time. More pipeline means more optionality. More optionality means less risk. This is the architecture of the partnership strategy most founders are running, and it produces exactly the outcome Greylock decided to protect itself against: a portfolio of relationships where you are not the most important partner to anyone, and therefore the one whose call gets returned last when the environment gets difficult.

The Cost of the Wide Portfolio

Partnership breadth produces the appearance of a strong network and the reality of a weak one. The founder running eight active strategic partnerships has distributed their relational attention across eight relationships, none of which has received enough investment to develop the depth that makes a partnership perform under pressure. Their partners know this. Not consciously, but operationally. The partner who is one of eight on a list develops an implicit reading of their priority status, and that reading governs how they respond when the partnership requires something inconvenient from them, a warm introduction that requires a real relationship, a deal term adjustment that requires goodwill, a conversation about underperformance that requires trust.

Those moments are where the partnership's actual value is realized or lost. And in every one of them, depth is the resource being drawn on. A shallow partnership has no reserves for the withdrawals that real work requires. What remains is a relationship that looks like a partnership from the outside, that shows up correctly in a slide deck under "Strategic Partners," and that functionally produces nothing more durable than a mutual agreement to speak well of each other at conferences.

The cost of this pattern is not always visible immediately. In the first six months, the wide pipeline looks like momentum. Introductions are being made, calls are being scheduled, partnerships are being announced. The announcement is usually where the relationship peaks. What follows is a gradual recalibration on both sides as the work reveals what the enthusiasm concealed: that neither party invested enough in the relationship to know what the other one actually needs, that the stated alignment was a match on aspiration rather than on operating style, and that the accountability structure required to maintain any partnership through difficulty was never built because there was not enough depth to build it on.

Greylock understands this. They have been investing since 1965 and they have watched enough partnerships between founders and capital deteriorate to know precisely which variable predicts the ones that hold. The variable is not sophistication of the agreement. The variable is not quality of the initial match. The variable is whether the relationship contains enough depth that both parties treat the other as consequential when consequences are on the line.

The Relationship as the Investment

The founders who build the most durable partnership portfolios in 2026 are running a version of the Greylock model whether they know it or not. They have fewer active partnerships than their peers. They invested in specific relationships months before any agreement existed. They can tell you, specifically, how each partner behaves when a deal stalls, what they prioritize when their attention is divided, and what they have done historically in the first weeks after a shared commitment begins to slip. They know these things because they spent the time to learn them, in conversations that had no commercial urgency, before any term was on the table.

The selection process that produces this knowledge is not compatible with the volume approach. You cannot deeply vet eight partners at the same time that you are maintaining eight active partnerships while also running a company. The founder who tries to do all of it produces a version of each task that is insufficient for the purpose: vetting that is too shallow to surface the behavioral signals that predict execution quality, and maintenance that is too infrequent to keep any single relationship above the threshold at which it can do real work.

What Greylock made explicit in this fundraise announcement is a resource allocation decision that most founders never articulate clearly enough to act on: depth requires scarcity, and scarcity requires choosing. The firm that "could have easily raised a multiple" of $1.5 billion and chose not to is naming the cost of scale at the expense of importance. That cost shows up in portfolio performance over a decade, the same way it shows up in a founder's partnership portfolio over two years, in the number of relationships that were active but never consequential.

The most important partner in any founder's network is not the most prominent one or the most recently acquired one. It is the one who treats that founder as consequential when the situation is inconvenient, who picks up the call during a difficult quarter because the relationship has the depth to make that reflexive rather than calculated. Building that relationship requires choosing it, investing in it before the agreement exists, and accepting that the cost of doing it well is the partnership volume you will not pursue because you are already committed somewhere else.

Platforms like onSpark are built around exactly this selection discipline, matching founders to the specific relationships worth depth rather than building pipelines for breadth. But the structural decision behind those matches, the willingness to invest fully in fewer partnerships rather than lightly in many, belongs to the founder. Greylock just made it explicitly in public. The founders who have been doing it implicitly for years already know what the portfolio looks like when it works.