Tuesday, April 16, 2013

Your Product Has No Curator

It's tempting and romantic to think a team or organization will naturally coalesce around a product. It's also very dangerous to think it actually works. It's a product & organizational smell.

You keep using that word. I do not think it means what you think it means.

Under the banner of 'enlightened management', organizations and teams may decide that there will be no formal management; employees choose what to work on and no one has the authority to say 'yes' or 'no'. Valve software is often used as model to describe this type of style. Google and Facebook are also used as examples of 'engineer lead' cultures (we'll come back to Google and Facebook at the end).

Most people who say 'enlightened management' are referring to work done by Abraham H. Maslow and his book, 'Maslow on Management'. The problem is that most people either didn't read his work, cherry picked parts of it or just refer to it via hearsay. People like the part where employees are encouraged to be self directed and self-actualized, but they skip the fact that Maslow never said management was unnecessary.  For example, many proponents cite:

  • Assume that everyone is to be trusted. Tolerance for mistakes is essential for creativity to flourish and risk-taking.
  • Assume that everyone is to be informed as completely as possible of as many facts and truths as possible, of everything relevant situation.
  • Assume that everyone will have the same ultimate managerial objectives and will identify with them no matter where they are in the organization or in the hierarchy.
...but what is skimmed over is usually the parts about:
  • Clear objectives
  • Clear goals  
  • Clear decision policies
Without the above, things just don't happen or if they do happen the end result is less powerful, unfocused and unpolished. Two great, public examples where a lack of curation created problems are from Valve and 37Signals.

Valve does great work but when almost 10 years pass between the release of a game and the release of an add-on (not even a brand new release), customers get really upset and press leaks abound.

37Signals is and was in a much better position. As Ryan Singer describes: "projects simply were not getting shipped." Ryan is a very talented product designer and 37Signals has really smart people. That is why they quickly realized that: "Somebody has to actually pull the people together, and get all the different roles to connect with each other." In order to ship: "...somebody has to make that happen."

What About the Ants?

Here's a great video about how ant colonies work. Ant colonies are extremely efficient and yet have no 'management'; ants decided for themselves what to do....or do they....? An important part of the video is how status is communicated within the organization.  Ants share status via physical contact and since they are piled on top of each other, colony status can be instantly be communicated to almost every ant.

Getting back to Maslow, ants fit perfectly into those first 3 points mentioned above, but what about the objectives and goals? This is where ants and humans diverge. Ants are hard wired with colony objects (see video). There is no question; there is no deliberation. The how, why and when of what needs to be done never changes for them. For people this is not the case.

Engineering Lead?

Interestingly, I find that many of the people who point to Google & Facebook as examples of 'enlightened management' working... never worked there themselves. Google and Facebook are large organizations and a lot goes on there. Also consider that Facebook and Google only have a few revenue streams compared to all the other products they offer. Does anyone really think that the engineers on the Adwords can do whatever they want, whenever they want...that a billion dollar product is just organically lead by the engineering team? (hint it's not) For a time (I'm not sure if this is still true) the money making products were in an entirely different building with different management styles than Google's other, commoditized products.

The other important factor with large organizations is that they can afford waste. They can have 2-3 engineers experiment for several weeks and end up with nothing. In the startup world - this could sink your organization.

Herr Direktor

This is why teams, organizations and products must have someone who is trusted to curate objectives, goals and decision policies.  The trick is how to go about this; how to say 'no' and guide a team. Many people have written well on this topic and this blog will feature more posts on doing this. The only thing this article will add is that no one can argue with data. If the results of an A/B test say version #1 had 20% more conversions or all 100 interviewed customers said they had trouble dong 'X'....well...who can contest that?