Nate Clark's Musings

A few years ago I read a great book summarizing the scientific literature on productivity. Summarizing that summary below for my future self (I forget things) and in case anyone else finds it useful.

The best tips for being productive according to science:

1. Make a decision. your brain is hard-wired to crave control. If you’re faced with a task you don’t want to do, make some decision on the task to trick yourself into doing it. E.g., if you have to write something, decide to do the conclusion or make the figures first. By making the decision you will subconsciously enjoy the task more. The Marine Corp calls this “a bias toward action”.
   
   
2. Link decisions to the bigger picture to trigger self-motivation. Ask yourself why you’re performing a task and connect it with something important to you. E.g., I’m writing this blog post partially to help improve my friends’ productivity, which should improve their lives and sense of accomplishment; or I’m working on this software bug to make a living to take care of my family.
   
   
3. Work on the right projects. This may sound obvious, but the most effective work is work you don’t do. One particularly effective method for sorting out the wheat from the chaff is developing mental models and evaluating tasks against those models. E.g., I have a model in my head about how the foreign exchange (FX) markets operate. A project proposal at work to port some of our equities trading algorithms to FX didn’t make sense against that model because clearing a settlement in FX works completely differently. This happened to be accurate and not implementing saved us time and effort. Mental models provide a framework for directing our focus in the proper directions, helping people avoid distraction by the constant flood of email and other data.
   
   Research has shown that top performers actually work on fewer projects than average, but they disproportionately work on early-stage projects that give them opportunities the develop new mental models. Top performers are also constantly evolving their mental models by bouncing ideas off others and imagining what-if scenarios.
   
   
4. Set two types of goals: ambitious stretch goals, and simple straightforward ones you know how to accomplish. There’s a lot of evidence that having goals in general improves productivity. The straightforward ones, or SMART goals, provide you concrete targets to hit which sparks motivation in your brain. The stretch goals are necessary to spark creativity and force you out of the myopathy of considering only what you know exactly how to do.
   
   For example, a stretch goal might be to reduce the time from commit in Caprona to production deployment to 1 hour. I have no idea how to do that, so a first SMART goal along that path would be to create a framework to collect statistics from git, Jira, and TeamCity on how long each step in the deployment process takes that I will complete by the end of the week.
   
   
5. Become an idea broker. Across a few different studies, researchers found the most “creative” science papers and managers had a gift for combining well-known ideas from different fields or departments in new ways. Coming up with something out of the blue is very rare. By exposing yourself to different ways of thinking or solving problems, you’re more likely to see innovative ways to solve problems. Go to coredev tech talks. Talk to other dev teams and read their status updates. Read about what happens at other tech companies. Read books. This will make you more innovative.
   
   One hack to stimulate creativity/innovation: tension (e.g., a deadline) or some type of disturbance (e.g., switch teams). Changes to status quo or external pressure are often great for jolting one out of creative ruts.
   
   
6. The key to team productivity is ‘psychological safety’ meaning everyone feels free to speak up without fear of being embarrassed, immediately shot down, or punished. Effective teams work to ensure that everyone contributes to conversations – good leaders will call on people who are not talking as much – and team members must be empathetic to realize when someone may be upset by something they might say. Research has found that teams with more women are generally more effective due to increased empathy, which creates more safety.
   
   Psychological safety is virtually the only thing that correlates well with team productivity. Remote/in-person, diversity, individual talent level, agile/not-agile, and everything else you can think of showed much less correlation with team success.
   
   
7. Build a culture of commitment. That is, try to create an environment that emphasizes teamwork, building people up, and making employees proud and fulfilled above, say, profit. A couple Stanford profs studied hundreds of Silicon Valley startups over a 15-year period. The company cultures were grouped into 5 categories: the “star model” where only people from elite universities were hired and given broad autonomy, an “engineering model” where there weren’t as many individual stars but engineering groups held sway, firms built around autocracies where all decisions flow from an all-powerful CEO, bureaucracies with scores of middle management, and “commitment” firms. (Yes, these categorizations are super-fuzzy and unsatisfying). Star-model companies were some of the biggest winners (e.g., Google), but also were most likely to go out of business. Commitment culture firms outperformed the other styles in virtually every way: most profitable, leanest (fewest middle managers), most likely to go public., fastest product delivery rate, etc. When people feel a sense of commitment to their team, they are more likely to work harder and feel psychological safety.