Back to blog
Ops Operating System

Blocked on a Person: Measuring the Waiting in Your Delivery

5 min read
blocked on a personmeasuring blockersdelivery metricscycle time blockerswaiting time

The short version

  • "Blocked" on a ticket almost always means "asked someone, no reply yet". That is a different problem from a technical dependency.
  • Split the blocked status into two: blocked on a person, and blocked on work. They have completely different remedies.
  • Add two fields: who we are waiting on, and since when. Those two make the invisible cost visible in a week.
  • Once you can see it, most of it turns out to be waiting for information rather than for a decision.

Every delivery process has a blocked status and almost none of them distinguish between the two entirely different things it is used for. That conflation is why blocked items sit for days without anyone being able to say what would unblock them.

"Ticket says blocked. It means 'asked, no reply yet'."

What "blocked" usually means

Ask someone why an item is blocked and the answer is usually a sentence about a person: waiting for Ana to confirm the schema, waiting on a review, waiting to hear whether we include the export.

That is not a dependency in the engineering sense. Nothing is technically preventing progress; a message is unanswered. But it sits in the same column as "waiting for the vendor to ship an API", which genuinely cannot be accelerated, and the two get treated identically in every standup and every report.

Split the status in two

Status Means Remedy
Blocked on a person A question or review is unanswered Reduce answer latency, or route elsewhere
Blocked on work Another piece of work must finish first Sequencing, or accept the dependency

The split takes ten minutes to configure and immediately changes the conversation, because the first row is actionable this week and the second usually is not. It also stops a genuinely unavoidable dependency providing cover for a pile of unanswered questions.

The two fields that make it visible

On anything blocked on a person, require two pieces of information: who we are waiting on, and since when.

Those two turn an invisible cost into a number. "Fourteen items blocked on a person, average age 2.4 days, nine of them waiting on two people" is a sentence that produces action. "We have some blockers" is not.

Expect mild resistance, because naming a colleague feels accusatory. Frame it explicitly as a measure of the system rather than the person: if nine items are waiting on one individual, that is a routing and latency finding, and it is usually evidence that the individual is overloaded rather than unresponsive, as in reducing dependency on the one person everyone asks.

What a week of data shows

Three findings are nearly universal.

The waits cluster on two or three people, usually the most senior and the ones holding the most context. The average age is longer than anyone guessed, because each individual wait felt like "just since yesterday". And the majority of the waits are for information rather than for a decision, which is the finding that matters most, because information does not require the person to be awake.

That last category is where StandIn changes the number. Each person confirms a ninety-second brief at the end of their day, and while they are off, their StandIn answers questions from it in their words, with a source under every answer, and never guesses. An item blocked on knowing where something stands stops being blocked at all, and an item blocked on a genuine decision is correctly identified as such within seconds rather than after two days of silence.

Re-measure after a month. The count of items blocked on a person and their average age are the two numbers that tell you whether anything actually improved. See how StandIn works.

Acting on it

Four actions, in order of speed.

Set a review commitment: reviews within one working day, in the reviewer's own zone. Publish ownership so questions stop routing by guesswork. Give the two people who absorb the most waits a way to answer without being interrupted. And for anything blocked on a person for more than two days, require an explicit decision: escalate it, route it elsewhere, or proceed on a stated assumption and record the assumption.

That last rule is the cheapest and the most neglected. An item waiting silently for four days is worse than one that proceeded on a written assumption, because the assumption can be checked and the silence cannot.

Common Questions

What does blocked on a person mean?

It means a question, review, or decision request is unanswered, so work has stopped even though nothing technical is preventing it. It is distinct from a dependency on another piece of work and needs a different remedy.

Is it fair to name who we are waiting on?

Yes, if it is framed as a measure of the system. Nine items waiting on one person is information about overload and routing, and hiding it protects nobody: the person is already carrying the load, just invisibly.

How long is too long to be blocked on a person?

Two working days, after which something explicit should happen: escalate, route elsewhere, or proceed on a recorded assumption. The specific threshold matters less than having one.

What should we measure?

The count of items blocked on a person and their average age, reviewed weekly. Both are simple, both move when you change something real, and neither can be improved by working harder.

When you're off, your StandIn is on.

It answers your teammates' questions from work you've already done, in your words, with a source under every answer.

You might also like