The paper ships at 11 p.m. The press does not wait for you to finish the sidebar, and it does not care whether your web publishing script is cooperative tonight. You have a window, and when it closes the trucks leave the loading dock without you. Standing in a composing room in Philadelphia in 1997, I began to understand deadlines through the people who depended on them.
I had joined Knight Ridder two years earlier as a software engineer, working on the web side of Philadelphia's newspapers. We were hand-building templates, publishing scripts, story converters and photo pipelines. We did not have a ready-made content management system that did the job for us. You wrote the thing, you ran the thing, and if it broke you fixed it before the copy desk noticed.
A deadline is a promise to strangers
In software, “deadline” often means the date we said we would ship, followed by a discussion about whether we can move it. In a newspaper, the deadline is a promise made by the building to the city. Trucks have to leave. Carriers have to reach their routes before dawn. Readers expect to find the paper on the porch. If the paper is late, a quarter million people know within an hour, and some of them call the circulation desk to tell you so. They will not see your Gantt chart or care about your explanations.
When the newspaper website went down at ten at night, we were already awake. The print deadline was coming, and the web had to be ready to publish the same stories within minutes of the press run. Waiting until the next morning would miss the work we were there to do. The rhythm of the building was the rhythm of the work.
That is still the first thing I want to understand when someone gives me a launch date: what happens outside the project on that date? Another organization may be sending people our way. An event may be bringing an audience together. Or the date may be an internal target chosen to organize the work. Each deserves attention, but the consequences of delay differ. I cannot decide what to protect until I understand those consequences.
Make delivery ordinary
A newspaper is the most relentless release cycle I have worked inside. Every day, seven days a week, holidays included. The next edition is already in motion before the current one is on the truck. Whatever review or planning needs doing has to fit around that continuing obligation.
It is easy to treat a software release as a heroic push, to talk about the team rallying for the final effort. In a newsroom, it is just Tuesday. The discipline has to live in the infrastructure and the habits. If every release needs a heroic push, we have built the wrong system.
Coming from university computer science into a newsroom, I noticed how reporters and editors handled unfinished work. A reporter could file a story with a missing fact flagged for the copy desk. An editor would chase the fact or cut the line, and the paper would go. Holding the entire edition while one sentence waited for perfection would serve readers badly. So would printing a claim we did not trust.
That taught me to separate the work that must be right for this edition from the work that can take another form or wait for another opportunity. A correction tomorrow cannot give back the chance to inform someone today.
Iterations and later releases are useful. I want teams to use them. I also want us to decide whether the thing in front of us can serve someone tonight, instead of treating the possibility of another release as an answer to that question.
Protect the reason people are coming
In January, Big Brothers Big Sisters of America had an opportunity to reach people through the front page of Yahoo. The organization was launching its new website as part of its centennial celebration, and the exposure was tied to a particular day. When the site ran into trouble before launch, I joined a call to help the team get it working.
Robin Palley, the organization's vice president of marketing and communications, described the circumstances in a letter this month: the launch time was fixed, with no opportunity to reschedule. The site went live, and Palley reported that the Yahoo feature generated 11,000 new volunteers.1 A working site the following day would still have been useful, but it would not have recovered the same opportunity.
For a site inviting people to volunteer, getting a page onto the web is only part of launching. An interested visitor needs a usable way to take the next step. A page that explains the mission beautifully but leaves that person stranded has missed the reason we invited them.
I would follow that visitor's intended action from arrival to completion, including what happens if they make a mistake or need help. An additional page about the organization's history might be valuable without being necessary on the first day. A clear explanation of what happens after submitting the volunteer form might be essential. Cutting both because they are “content” would miss the difference.
Protecting the personal information that a form collects is part of making it work. So is making the form understandable and usable. If we cannot meet those requirements, the people arranging the announcement need to know while they still have time to change the message, narrow the promise or make another arrangement. “Engineering is working hard” gives them very little to act on. I want to explain which visitor action is at risk, what remains uncertain and when we can make the next decision.
I am glad I could help that night. I want the preparation around a launch to make late-night rescues less necessary. People sometimes do valuable work under difficult circumstances; their willingness to help should not become the staffing plan.
What I carry into web work
The web changed the daily deadline into a stream of smaller ones. Stories appear through the day, receive updates and reach readers through alerts and email lists. The ability to update is valuable. Readers still experience a story partly through when it reaches them.
The engineers I worked with in the newsroom understood this because they sat next to editors who treated time as a physical material. You had the time you had. You made the call, or the building made it for you. The web teams I have worked with since have been better or worse according to how seriously they took that obligation.
The tools change. What stays with me is a way of looking at the clock. When a date can move, moving it may be the responsible choice. When it represents a commitment to people outside the team, I want us to arrange our work so the trucks leave with what those people need. I know which kind of team I want to build: one that makes keeping that promise an ordinary part of its work.
I think about the composing room sometimes, the sound of the press starting and the way the room emptied at 11 o'clock because the day's promise was kept. It was a quiet moment. That is the part I try to take with me.
Footnotes
-
Robin Palley, letter of reference for Rajiv Pant, March 8, 2004, published with permission. ↩