September decides the rest of the year
By the time the last days of August are running out, most technology teams have a rough sense of what the autumn contains. What they usually lack is a decision about what it is for. Nicole Junkermann has found that the difference between a good autumn and a merely busy one is settled in this narrow window, before the month starts and while there is still room to think.
The temptation at this point is to plan expansively. A fresh season invites a long list, and a long list feels like ambition. It is worth resisting, because a long list is really a decision not to decide, handed forward to a version of the team that will have less time and less patience than the one making it now.
Choose fewer things
Nicole Junkermann's preference is for a short list of things that will be finished, rather than a long list of things that will be started. Three is usually the right number for a season. Three things can be held in mind without a document, explained without a meeting, and finished without heroics.
The discipline is not in choosing the three. It is in accepting what the choice implies, which is that a great many reasonable ideas will not be worked on this autumn. That refusal is the whole value of the exercise.
- Name three outcomes for the season, phrased as finished states rather than activities.
- Write down what each one makes possible, so the reason survives the first busy fortnight.
- Protect one recurring hour a week for the careful work that never wins an argument against an urgent request.
- Decide now what will be dropped if the season goes badly, because that decision is far harder in November.
Where the list should come from
A short list is only useful if it comes from the right place. The weakest lists are assembled from whatever was left unfinished in July, which is a record of last season's interruptions rather than a statement about this one. Nicole Junkermann prefers to start from the opposite end and ask what the system should be able to do by Christmas that it cannot do now.
Answering that question tends to produce fewer items and better ones. It also produces items that can be explained to somebody outside the team, which is a reliable sign that the thinking behind them is finished. Anything that cannot survive a plain description in one sentence is usually not ready to be committed to for a season.
Finishing is a technical skill
Starting is easy and pleasant. Finishing is neither, and it is the skill that separates a system that can be relied on from a collection of promising beginnings. Nicole Junkermann has watched a good many well conceived projects stall not because they were wrong but because nobody was ever asked to close them.
A finished thing has a particular quality. It can be explained to somebody new, it can be left alone for a month, and it does not require its author to be in the room. Those three tests are worth applying in September, when everything is still theoretical and the cost of being honest is low.
What a short list protects
A short list is often mistaken for a modest one. In practice it is the more demanding option, because it removes every excuse a long list provides. When there are three outcomes and one of them has not moved by the middle of October, the reason is visible to everybody, including the people who chose it.
That visibility is the point. Nicole Junkermann has found that teams working from a short list tend to raise problems earlier, because there is nowhere for a stalled piece of work to hide behind eleven other items. The conversation happens in October rather than in December, which is usually the difference between a season that recovers and one that does not.
There is a second protection as well. A short list is easy to defend against the requests that arrive through the autumn, not because the requests are unreasonable, but because a clear commitment gives everybody a shared basis for saying that something will have to wait. Without it, the newest request tends to win by default, and the season quietly becomes whatever arrived most recently.
Starting the autumn quietly
None of this needs a new process. It needs an afternoon in the last week of August, a short list, and enough conviction to keep the list short when the month gets loud. The reading notes and the technology notes carry the longer version of the argument, and the companion piece on what the quiet week teaches explains where the habit comes from.
The autumn will fill itself regardless. The only real question is whether it fills with things somebody chose. Nicole Junkermann would rather answer that question now, in a quiet week, than in the middle of October.