Plan or build?

July 25, 2012productivity

I met with Jesse Nowlin a few days ago, a great founder who’s really hustling. We spent some time discussing his idea and I shared some of my experiences with Buffer.

When we’d almost finished our 30 minute meeting, Jesse said “I have one last question”. He asked me the following:

"Plan vs build. Where do you stand?"

I thought it was a fantastic question. Clearly it’s not a binary choice, yet I think it’s also good to ponder which of the two you should focus on.

What does it mean to plan?

I believe planning is an essential part of making progress. I think there’s no way you could dismiss it. If you have a startup idea, or if you have an idea for a new feature, it would not be productive to start building without spending some time deciding what work is involved, how the idea should be approached and some of the milestones being aimed for.

That said, from my experience of building startups I now think the focus should be firmly on building over planning. I think a good ratio of building to planning would be 95/5 or even greater.

The daily standup

As an example, a daily standup has become a common and important ritual for teams. One of the widespread principles of a daily standup is a 15-minute duration:

"Keep the meeting to 15 minutes or less. Minds wander and focus is lost if the meeting continues beyond this time frame."

If you think about the ratio 15 minutes equates to in an 8 hour day, it is around 3%. I think this is perfect.

Brainstorms at Buffer

At Buffer, we wouldn’t get very far without the time we spend planning. Often we say to each other:

"can I spend a few minutes brainstorming what might be best for me to work on next?"

Other times we’ll brainstorm the next steps for a specific feature, or thrash out the technical details of how something should be implemented.

However, we’ve found that we have formed a culture with a focus on acting with incomplete information, and a bias for progress over discussion. When we brainstorm, after around 20 minutes one or more of us in the discussion will feel itchy to move on from the brainstorming and get back to building.

What does it mean to build?

I’ve come across some seriously smart people who decide they want to build a startup. One of the things I see people struggling with the most is getting out of the planning phase.

To start building takes real guts. To actually put something out there, to ship something, that’s scary. It’s terrifying because it then becomes real. All the plans are tested. However, we all know that “shipping” is what we must do to achieve great things.

"What you do for a living is not be creative, what you do is ship." - Seth Godin

You can’t learn by planning

A quote I keep coming back to, and find myself repeating to many founders I meet, is the following by Matt Mullenweg:

"Usage is like oxygen for ideas. You can never fully anticipate how an audience is going to react to something youve created until its out there."

This is something we try to embrace at Buffer. I truly believe you can’t learn a lot by planning. You’ve got to ship it and see the results.

The surprise of data

Until you start shipping regularly and thinking about everything in terms of how quickly you can get usage and test your assumptions, it’s easy to imagine that things will often go to plan. The crazy part is, in my experience, the more likely case is that things won’t go to plan.

Here are two examples of data which surprised us recently:

• In a new version of the Buffer browser extension popup, for which we conducted user testing, we found a key feature was completely misunderstood by 90% of users.
• We built an MVP of a new weekly email feature, aimed at being sent to all users. We manually calculated results which would later be calculated by an algorithm and found that the results were only good enough for us to email 40% of users.

What can we do right now?

In both of the above examples, we were lucky to have created a culture of doing things lean. What this means is that we don’t feel invested in something when it fails, and we can quickly make decisions about adjustments to work towards great results.

As a result of doing more of these quick lean validation activities, we are increasingly asking ourselves "what can we do right now?" whenever we plan a new feature or significant change which has untested assumptions.

How do you feel about planning vs building? What is the approach you take with balancing the two activities?

Photo credit: Dave B