Getting the title to match the work

Almost everything written about promotion packets treats them as a submission — a document assembled at the end of a cycle to argue a case. The people who get the most out of them use them the other way round, as a running check on whether the work they are doing is the work the next level is described by. That reversal is also what makes the document possible to write at all, because the evidence it needs stops being available surprisingly quickly.

The situation people are usually in when they look this up is a specific one: the work has clearly changed, everyone says so informally, and the formal case for it now has to be assembled out of a year that was not documented.

What the packet is really for#

Will Larson’s guidance on staff promotion packets contains the sentence that changes how the whole exercise is worth approaching: you will spend more time relying on it as a guide than as a formal artefact for official review, so optimise for the former. A document written for a committee is written once, at the end, in the most stressful possible week. A document written for yourself is open all year, and its blank sections are the most useful thing in it.

That matters because a packet organised against your company’s actual framework will have blank sections, and they are informative. Published frameworks of this kind — Dropbox’s is the usual public reference — describe each level in terms of scope, collaborative reach and the ways impact is achieved, assessed across results, direction, talent and culture alongside role-specific craft. Four of those five headings are not about shipping. A person who has been heads-down for three quarters can see that on the page in about a minute, which is a year earlier than they would find it out in a review.

What goes in it#

The contents Larson lists are worth taking literally, because each one covers a category of work that vanishes from the record by default.

Projects at the level you are arguing for, each with what was actually accomplished, how complex it was, and links to the supporting material. The links do more work than the prose: a document written at the time is evidence, and a summary written afterwards is a claim.

Changes to how the organisation works — the improvements whose effect is much larger than the work that produced them. These are the least self-evident entries and the ones most worth writing carefully, because their value is invisible to anyone who did not experience the situation beforehand.

People you developed, and what they went on to do. Naming their accomplishments rather than your mentoring is the difference between an assertion and a demonstration.

Glue work — the coordination, unblocking and unowned tasks that hold a project together — with what it made possible. This category is where the largest amount of genuinely senior work goes unrecorded, precisely because none of it produces an artefact with your name on it.

Advocates: the teams and leaders who will speak for the work, with the specific value each of them saw. Not a list of people who like you — a list of people who can describe a concrete thing that was better because of you.

Gaps: the concerns about your skills or behaviour, and how you would address them. This is the section people are most tempted to leave out and the one that makes the rest credible. Every committee has an objection; the only question is whether it is raised in a document you wrote, with your answer attached, or in a room you are not in.

Why it has to be gathered early#

Larson’s advice is to start the first draft long before you think you are likely to be promoted, in the way one would keep a running record of one’s own work, on the grounds that promotions at this level are built over quarters, halves and years. The reason that is a practical instruction rather than a counsel of perfection is that the individual pieces of evidence decay on different, mostly short, timescales.

The measurement that made a project matter is often available for a fortnight. Dashboards roll up, retention windows expire, and the before-and-after that was obvious in the moment becomes an assertion. The colleague whose team you unblocked changes team. The document that shows what the design looked like before your objection is superseded, and only the current version is easy to find. None of these is dramatic and together they account for most of what is missing from a packet written at the end.

What replaces all of it is unglamorous: a file, one dated line per entry, written the same week the thing happened, with a link and the number as it stood that day. Three minutes, perhaps twice a month. It is worth being precise about how small the habit is, because the way people fail at it is by attempting something larger.

Where it goes wrong#

The commonest defect is a packet made of activity rather than of consequence, and it is easy to produce while being careful and honest.

Every line begins with a verb about you. Led the migration. Drove the adoption of the new client. Coordinated across three teams. All of it happened; none of it is checkable; and a committee reading twenty such lines learns only that the year was busy — which was never in doubt and is not what the decision turns on. The repair is mechanical, one line at a time: state what is different now, and put yourself in the second half of the sentence. The migration finished a quarter early and removed the on-call rotation that came with the old system; you led it. The first clause can be verified by someone who was not there, which is the entire function of the document.

The related trap is reaching for numbers that will not survive contact. A figure attached to a change you cannot isolate is worse than no figure, because the one part of your packet that gets audited is the part with a percentage in it. The same discipline that applies to reporting on a team’s delivery applies to reporting on yourself — the difference between a measurement and a number that flatters the person reporting it is exactly the difference a committee is trained to look for.

The failure that costs the most, though, is passive: leaving the writing to your manager without supplying anything. It is not a matter of their goodwill. A manager writing from memory reconstructs what was visible from where they sit, which over-represents the work adjacent to them and under-represents precisely the glue work and the prevented problems that the level is defined by. The packet then argues honestly for a smaller version of your year, and the committee decides correctly on the evidence in front of it.

One thing a packet cannot do is create the scope it describes. If the work has not changed shape, no amount of documentation persuades anybody, and the questions worth answering first are elsewhere: whether the decisions are reaching you earlier than they used to, and whether the seat you are arguing for exists in this organisation at all.