Summary: "People often make the mistake of focusing on what should be done while neglecting the more important question of who should be given the responsibility for determining what should be done. That's backward. When you know what you need in a person to do the job well and you know what the person you're putting into it is like, you can pretty well visualize how things will go." — Ray Dalio

The principle

Dalio frames decision-quality as a function of who is making the decision first, and what gets decided second. Most managers do this in the reverse order: build a plan, then look around for someone to run it. He argues that order produces fragile execution — because once you've defined the work, you've also implicitly defined which kind of judgment, ownership, and adaptability the work requires, and any mismatch downstream is now baked into the plan.

The diagnostic move is to visualize:

  1. Visualize what the job actually requires (judgment, taste, adaptability, domain knowledge, tolerance for ambiguity).
  2. Visualize the specific person you're considering placing into it.
  3. The gap between (1) and (2) is what's going to happen.

Where the community pushed back — and why the nuance matters

The post drew a sharp set of replies that improved on the principle rather than refuted it. Three threads stand out:

  • Johnathan Johannes — "It's not 'who over what,' but getting both right at the same time. The right person with the wrong strategy will still fail. And the right strategy in the hands of the wrong person will stall." The real leadership work is alignment — clarity of direction matched with capability, ownership, accountability.
  • Jim Bascom (via Johannes) — once an organization is in motion, the who absolutely becomes decisive. Judgment, adaptability, and ownership start to outweigh the original plan. "The who doesn't just execute the what. The who continuously reshapes it." The real leverage isn't choosing between them — it's selecting people who can evolve the what in real time as reality unfolds.
  • Grayrock Global — most allocation errors trace back to a person mismatch, not a strategy flaw. Hiring for accountability first and task definition second is how the best investment teams operate.

A tighter restatement of the synthesized principle: the who is upstream of the what, but the what still has to exist — and the highest-leverage move is putting decision-rights with the person whose judgment can keep the what coherent under change.

Why this matters for architecture and engineering leadership

The principle reads as a leadership cliché but it has a sharp edge in technical organizations:

  • Architecture decisions are recurring judgment calls under uncertainty. Putting them with the wrong person — even with a great spec — produces a system that fails on the seams the spec didn't see. This is exactly the failure mode Zang Jing Ge is trying to prevent by codifying technical excellence, communication mastery, and leadership power as architect attributes.
  • Delivery work runs on judgment about scope, sequencing, and trade-offs that no spec captures fully. Role of a Delivery Architect is partly a "who" definition.
  • Maker–multiplier balance — Maker vs Multiplier is the same observation in a different frame: as an engineer's career grows, who is doing the work and who is shaping how others do work becomes the dominant performance lever.
  • Culture and accountability — Symptoms of a broken product culture is the inverse picture: when the what is endlessly redebated because nobody trusts the who, the organization is failing the principle from both directions.

Where the principle breaks (or needs a counterweight)

Two counterweights worth holding in tension:

  1. Process matters when the who pool is small. In a constrained team, you can't always staff to the ideal who. The discipline then is to define the what tightly enough that a good-but-not-perfect who can still execute it — which is exactly what Encoding Team Standards is for.
  2. Judgment-first hiring is a slow loop. Selecting on who requires evaluating judgment, which takes more signal than evaluating skills. Many organizations don't have the patience or interviewing depth for that. The principle is correct; it's also expensive to apply.