Why portfolio construction advice runs out once you have closed

The observation

There is a large, well-made body of writing on venture portfolio construction. Almost none of it is addressed to a General Partner who has already closed a fund.

This is not a criticism of any of it. The material is good, and much of it is better than what we would write on the same subjects. The point is narrower and it is about the addressee: read piece by piece, the corpus consistently assumes a reader who is still choosing, still fundraising, or freshly closed and yet to deploy. A General Partner three years into a fund is not that reader, and there is very little written for them.

The pattern, and how to check it

The clearest way to see it is in how the material frames itself. The fund-formation and portfolio-planning content that ranks for these questions is organized around launching: how to plan a first fund, how to determine a fund size, how to form the entity, how to build a construction model before the raise. Portfolio construction is typically defined outright as the plan for what you will do, the number of investments, the average cheque, the target ownership.

Modelling material has the same shape for a different reason. A construction model is at its most valuable while assumptions are still being chosen, so writing about modelling naturally addresses the moment before the assumptions are locked.

This is an observation rather than a study, and it is offered as one. The way to check it is to take any guide on venture portfolio construction and ask a single question of it: does this help someone who cannot change their fund size, their investment period, or what they promised their LPs. Our reading is that the honest answer is almost always no.

Why the gap exists

There is a reasonable explanation, and it is not that anyone overlooked the audience. Pre-close is when the advice is most obviously actionable. Everything is still a choice, so a guide can tell a reader what to pick. It is also when the reader is most actively looking, because raising a first fund produces an urgent, searchable set of questions in a way that year three of deployment does not.

And the honest difficulty: once a fund has closed, most of the construction is fixed. Fund size is fixed by the partnership agreement. The investment period is fixed. Stage focus and industry scope are fixed by what was promised. A guide that tells a General Partner to choose a different fund size is useless to someone who cannot.

So the gap is real, and it exists because the post-close question is genuinely harder to answer than the pre-close one. Not because it is less important.

What the post-close question actually is

A General Partner inside a closed fund is not asking what to choose. They are asking a different question, one the pre-close corpus never has occasion to address: given what is now fixed, what is still moveable, and which of my remaining decisions are pulling against the commitments I have already made.

That question has real answers. Ownership targets can be revised. Portfolio count can be revised, and usually is, in practice if not on paper. Reserve deployment is a live decision every quarter. Board practice, lead practice, team size, and platform commitment are all adjustable. The next fund's construction is entirely open and is best informed by an accurate reading of this one.

What a General Partner in that position needs is not a plan. It is a reading: which of the things I declared still fit together, which have stopped fitting, and what does the misfit actually cost. That is a different instrument from a guide, and it is what this library and the Colibrí Architecture model are for.

Why this is the thesis

Every page in Architecture Intelligence sits on this observation. The eleven congruence checks are readings of a firm as it stands rather than advice on how to set one up. The reserve pages take a reserve percentage as already chosen and ask what it implies. The drift page exists only because a fund keeps moving after the plan was written.

The library's subject is a fund that has already been raised and is already being deployed, because that is where a General Partner spends nearly all of their time and where almost nothing has been written for them. If the observation on this page turns out to be wrong, the library has considerably less reason to exist, which is why it is stated plainly enough to be argued with.