Portfolio governance guide
Distributed organizations collect work through too many channels. Intake discipline makes work visible before prioritization turns it into a capacity decision.
Requests arrive as escalations, executive asks, roadmap assumptions, customer pressure, compliance needs, and informal priority claims. When everything enters as urgent, value, risk, dependency, and timing get blurred, and the organization loses the ability to tell a funded commitment from a hopeful one.
Separate intake from prioritization. Intake makes work visible: what is being asked, by whom, and against what evidence. Prioritization is the later, separate decision about whether it deserves capacity, when it should move, and what tradeoffs it creates.
Fewer hidden commitments, cleaner escalation paths, less duplicate work, better sequencing conversations, and stronger executive confidence that the portfolio list reflects what is actually happening rather than what was once approved.
A working model
Intake is not a form. It is a short sequence of gates that a request has to clear before it is allowed to compete for capacity as a real portfolio item.
Practical interventions
The first job is not a better form. It is stopping the organization from treating every request as if it were already a decision.
Capturing a request is clerical. Approving it is a decision. When the same conversation does both, logging demand quietly becomes committing to it, and the backlog fills with items no one remembers agreeing to.
A request without a named sponsor and a named decision owner should not be allowed to reach a prioritization conversation. Assigning ownership after the fact almost always means no one actually owns it.
Deferred, blocked, and rejected work needs the same record-keeping as approved work: what was asked, why it did not move, and what would need to change to revisit it. Without that ledger, the same request resurfaces every quarter as if it were new.
Hidden commitments, duplicate requests, sponsor claims without ownership, urgency that is really escalation pressure, and work entering delivery before it has enough context to evaluate.
How this plays out
In one enterprise environment moving through several years of rapid, acquisition-driven growth, technology investment requests arrived from a dozen directions at once — platform teams, security, finance, sales operations, and newly acquired business units each running their own informal priority list. At any point, several hundred initiatives were active or proposed, with no shared way to compare them or confirm which ones still had a sponsor.
The fix was not a new tool. It was a CIO/CFO-sponsored governance council with one intake standard, one decision log, and a qualification gate that every request had to clear before it was allowed to consume delivery capacity. Portfolio conversations moved from anecdotal updates to structured sequencing within months, without slowing the teams that were already shipping.
Avalara
The tremendous work and discipline he has brought to the Platform Services Team … a huge step forward for our billing processes as well as our company.
Where this breaks
When every request is flagged urgent, urgency stops meaning anything and the qualification step gets skipped under pressure.
Work enters delivery before intake happens at all, usually because a senior stakeholder asked verbally and no one wanted to slow them down.
A request clears intake with no durable owner, so it drifts through backlog reviews until someone finally asks who is accountable for it.
A form gets filled out, but no decision follows it, so the form becomes evidence of activity rather than a gate that actually filters anything.
Rejected or deferred work has no visible record, so it is re-argued from scratch every time the original requester raises it again.
Decision test
The request has a named owner, a clear problem, enough value and risk context to compare, and a known decision path.
The request may be real, but it still needs scope, sponsor clarity, dependency checks, business context, or evidence before it should consume portfolio capacity.
The organization can explain why the work is not moving now without losing the record, the rationale, or the ability to revisit it later.
Questions this raises
Only if capture and approval stay merged. Splitting them removes friction later, because fewer half-committed requests reach delivery only to stall.
Exceptions are fine. An exception with no name and no rationale is not an exception — it is a bypass that will get copied.
Intake makes a request visible and comparable. Prioritization is the separate decision about whether it earns capacity now, later, or not at all.
Continue