Victus Spiritus


When it comes to Startup Tasks, there can be only One

18 Jul 2010

Six months ago we tracked our startup's work tasks with a combination of lighthouse milestones, dozens of tickets and a metric ton of emails per day. In addition a Google task list, doc file, and a local text list kept track of my personal tasks and items I planned on working on. At the same time whenever I came across potentially useful documentation, I'd categorize and bookmark the site. I don't use any of those tools anymore*. I'll describe how and why that has helped me become more efficient, and more importantly brought clarity to my problem solving.

No one can accurately predict the future

We constantly get new relevant information that guides our decision making. What you perceive as necessary or useful now, may be irrelevant and a waste of resources a week from now or even tomorrow. The most important resource in a startup is your team's time and along with it the money to keep everyone working.

When your team is actively working towards a dead end, it's everyone's job to identify and communicate correction to assumptions and earlier data. While a startup can survive passing up an opportunity, sprinting in the wrong direction can send it off a cliff. Course correction with involved task management software requires changing priorities for task lists, milestones, emails, and miscellaneous documents that plan effort, and those activities all cost precious time.

Ducktyping Bureaucracy

If it looks and feels like bureaucratic overhead it is. And your startup should avoid it like the plague. Heavyweight ticket tracking and project management systems are a necessary evil for companies that need to support potentially millions of customers.  Once your startup has discovered a solid market fit automated ticket organization can help minimize repeat bug reports and explanations (DRY). Before your team identifies a scalable resource project tracking is overkill, and will cause you to miss vital early feedback. Prototypes are required to be just solid enough to generate measurable user response. The filthier the hack job, the faster a product fit can be found.

Concentration is King

Once you remove all the misdirection and murky obfuscation by ruthlessly purging task management lists, you'll find one vital thing to do today. That's your new to do list. Only when you complete that most important thing can you add on smaller tasks.

Only add one task at a time and complete it like a FIFO buffer. Bigger tasks can go on for days or even weeks. There's no need to continually waste time and effort juggling a list of one. If something else trumps it in priority, deal with the interrupt as a new dominant task. Avoid the decision making error of sunk costs when returning to uncompleted tasks. Some tasks are meant to die


*= old habits die hard, I continue to email myself potentially relevant docs and label them for later reading