Better Product Managers, and Product Management

Even when you’re surrounded by people who have fully bought in to the “lean” philosophy, it’s easy to fall into bad habits and start overbuilding things.

Lately we’ve set ourselves the following challenge: any task that will take more than a day to implement needs to have a written spec and estimates of how long it will take to complete.  Now, no one likes writing specs (our user stories are pretty spare and get fleshed out through comments in Basecamp) and we like doing estimates even less.  So we’re particularly motivated to look again at longer tasks and figure out how to do some version of them in less than a day.

(Devotees of GTD will recognize this as a variation on the ‘if you can complete a task in 2 minutes, just do it’ concept.)

Obviously, not all tasks fit into this mindset, but I’ve been pretty amazed by the amount of really useful one-day-or-less tasks we’ve been able to negotiate on, implement, and release.

I wanted to redesign our KISSinsights results page to incorporate pagination, “new responses since last login”, search, and filtering.  This would’ve required at least a couple days (and since it required multiple people collaborating, more than that in elapsed time).

Instead, we released search within results.

And then the ability to filter responses by people who chose a specific answer option.

The results page is now at least 25% more useful than it was 2 days ago, and no one had to write a spec or wait around for someone to free up.

It’s not just an engineering thing, either – I get these grand ideas and get all excited and write myself a task.  A few weeks ago, I created myself a task to create a series of marketing emails for my product to be sent out at weekly intervals.   It has been sitting there.  Why? Because it’s not a 1-day project – it’s a ‘wait until I have some quiet time to concentrate and do a bunch of writing, and then format all those emails’ project.   And we all know how often that happens.

So today I’m deleting that task.  I’m going to write one marketing email today.  Maybe I will write more later, but I will get one done today and cross it off and get it implemented.

What’s the biggest thing bugging you? I challenge you to find something you can do to make it better that you can complete today.  It may only be a tiny thing, but it will be better and more importantly it will be done.

When is a product task done?

A) When the requirements are complete?

B) When the code has been written?

C) When it has been deployed to production?

Answer: none of the above. A task isn’t done until you’ve validated that it accomplished your goal.

( You did have a reason for adding that feature, right?)

Here’s a platform-agnostic view of what we’re doing with our product backlog now:

• Write requirements for the task as usual.
• Add a hypothesis of how you think this task will improve your product.  Numeric metrics are  good but not always necessary.
  • example: Adding a guided First User Experience will increase visitor signup from 10% to 20%
  • example: Moving the “password reset” link will reduce the number of customer support emails we get
• Add when and how you will measure whether or not that improvement was realized.

and then tasks will move across 3 to-do lists:

That way, tasks don’t languish on a to-do list for ages – but they also don’t get marked as “done” until we are sure that they did – or didn’t – achieve what they were supposed to achieve.

(Ideas inspired by our team’s conversation with Eric Ries yesterday – it was a great nudge from him to move this knowledge from our heads to an actual, trackable system.)

I love Getting Things Done. Some of the cornerstones (like maintaining an empty inbox, for example) are a little too hard-core for me, but mostly it’s a philosophy that’s contagious in a good way.

One of my favorite former coworkers, a lead engineer, was also a fan of GTD. We always had a series of little tasks for each other – not fire-fighting emergency things, but “when you get a chance” items.  In order to not lose those items, we each had an online ta-da list that the other one could add things to.

Ta-da Lists are fast and easy – you get your own URL where you can create as many lists as you want (and yes, give permission for others to add items to your list). I maintain short-term lists while I’m in the midst of projects and an odds-and-ends list for home and one for work. The odds-and-ends lists are critical so you’re always one click away from remembering what useful thing you can do in that extra five minutes.

(more…)