Recognising 'good enough'

One of the trickiest aspects of my work-life is recognising when an output is ‘good enough’. I am by no means a perfectionist, though I do take a real sense of pride in doing good work. The problem lies in when I spend too much time trying to make the work good, when it would have been better to finish when it was good enough.

There are several areas in which this concept applies. Within shallow work, there is the temptation to polish an email until it is perfect. I do believe that there is a certain amount of effort (and at the very least a single proof-read) required to write a good email. The content should be clear, the message only as long as is needed, and any action items spelled out. However, most emails are fine to have a few grammatical errors, or to be more verbose than needed. Clarity and concision take a lot of effort, and if the content is time-sensitive or my day is busy, then an email sent sooner is better than an pulitzer-worthy email sent too late.

Within lab-work, I am eternally trying to strike the balance between the number and the quality of experiments. There is the organic chemist’s catchphrase “Garbage in, garbage out”: if the starting materials are impure, then the product will be too. Similarly, if the catalyst is impure, then the kinetics are not going to be reliable. This train of thought can be pervasive, questioning the results of everything: “how do I know whether this reaction didn’t work (or did!) because the material was contaminated or because the reaction will never work?”; “what if the purity of the reagent is a relevant variable?”. Soon, one could be tempted to spend weeks purifying every reagent to 99.999% analytical grade. But if the original question was “does water affect this reaction”, then that is the key variable to control. The first step is to test that hypothesis to a level that is ‘good enough’ to answer the question: try it with and without added water, using the same reagents.

Within deep work, when I am collaboratively writing a manuscript, I have to resist the temptation to write concise, stylish prose. As long as my intentions are clear, my collaborator(s) can add or subtract or argue any of the points. It is better to get them something early, so that they can have their input before I polish large amounts of writing that might be completely deleted in the final version. When I am editing someone’s work, I find it harder to make large suggestions when the draft is already nearing completion. The temptation is to polish what is there, rather than suggest a different order or a different argument. For a first draft, I actually think that ‘good enough’ is better for everyone, even though it feels like shirking work.

The Pareto principle, also known as the 80/20 rule, states that 80% of the outputs are achieved from just 20% of the inputs. Another way to think about it is that the difference between the ‘good enough’ 80% version and the ‘perfect’ 100% version is four times the amount of work. To polish that email will take much longer than to get it to the ‘good enough’ state. The amount of work achievable when ‘good enough’ is the goal is much higher than when aiming for perfect. Equally, when shallow work is becoming overwhelming, invoking ‘good enough’ might be the only way to stay on top.

The COVID-19 pandemic has now settled a bit, and is on its way to become endemic. Students are returning this semester,¹ and the research group has several new honours and PhD students. As a result, I am playing a tricky game of catch-up. The pandemic and resultant laboratory closure has meant that I have a lot of experiments planned - I had a lot of time to write-up what I had, which resulted in new ideas and new planned experiments. While trying to perform these experiments (with some working restrictions still in place), I am also attending to several new researchers and helping them with their work. This means that the highly time-blocked day that I got used to while working from home is now almost impossible to achieve. With four researchers vying for my time, every attempted deep-work block inevitably gets interrupted.

In order to have the time and energy to perform the high-quality outputs such as confirmatory experiments and polished drafts, I have to accept that some of the ‘shallow’ work will have to be just ‘good enough’. I cannot ignore the work, as part of my job includes activities such as replying to emails, helping honours students, and making sure the laboratory runs. But I can have a period of sending less polished emails, having more impromptu meetings and discussions (where I, for example, draw schemes by hand instead of having already prepared them), and also saying ‘no’ to more opportunities than I am used to.

Something I have been really struggling with is the temptation to optimise my productivity system. Having used Zettlr for my Zettelkasten (see this post for details), I spent a few hours last week looking at Obsidian. While some of the features look very useful, there is a time-cost incurred swapping from one piece of software to another. Ultimately, I worked out the minimal changes necessary to get both Zettlr and Obsidian to play nicely with the same vault.² This time felt very productive, and produced tangible outcomes. However, on reflection, I could have continued to use the system I had (Zettlr), and started the conversion in a few weeks after the new researchers are settled and I will have more uninterrupted time. Ultimately, my setup was ‘good enough’ before, and I spent a few hours polishing it with comparatively little gain.

All of this is to say, for the next few hectic weeks, I will be embracing ‘good enough’ where possible to maximise the amount of time I have for ‘great’.