The long game

What time have we got to develop an academic research project? What time have we got when we develop and launch an app, as an indie maker?

In my experience as a social scientist, academic projects leading to publications tend to last between 2 to 5 years. For an app (web app or else) launched by an indie maker, we often see projects that launch after less than 6 months, and the fate of the app is decided in less than a year.

For long, I have felt uncomfortable with both these time scales, because they do not help design long term, “life time” perspectives in terms of accomplishments. I believe I have found a better way. I am reprogramming my academic and dev work habits to reason at a 20 or 30 years horizon.

Academia: the pressure to publish tends to shorten and narrow down the research horizon

It is going to be hard to generalize about scientists, as academic fields are so different in their social organizations. Still, the vast majority of scientists are today under the intense pressure to publish frequently, in selective journals. This has for effect to reduce the time of research projects to a span which is compatible with short publication cycles.

Long term horizon science is of course possible and exists in many forms - just cut your long term research goal into smaller projects of 2 to 5 years each. Huge collaborative efforts lasting decades in physics are the proof that far sight exists. In all fields, some academics have the stamina to follow this fast paced publication cycle and to participate in longer-term institution building initiatives. This is how research institutes, new journals, new societies and other research networking initiatives get developed.

But in my case at least, I found that research projects that need more than a couple of years to spring and blossom were a non starter. The need to get the next grant, to renew the next contract, or to score the publication points for the next evaluation, ended up filling the time with plenty of interesting and fullfilling projects - also, which had the most chances to deliver strong results in a short time. There is always a timer running, and it is quite short. Maybe because I am in my early forties, but I now miss a sense of higher purpose. Happy and proud of each of the research pieces I have published, but how to make sense of this collection of projects in many directions? What has changed today is that I long for research projects which contribute to a bigger, long lasting “oeuvre”.

Indie makers: ship in 3 months

Indie makers developing web applications follow the principle that there is no point building an app for years, to have it ignored or rejected by the public when it launches. Fail fast, fail often. Launch your app as quickly as feasible (3 to 6 months are often heard as targets), so that it can pass the most important test: does it get any traction? If not, there is no market fit. You should kill the app and move on with the next project.

I used to follow this advice in the following way:

  • develop a half baked app, because I have not enough time and / or skills to do better
  • “launch it quickly” (meaning, the web app goes live)
  • waiting a couple of months, making few meager attempts at evolving the app, but getting zero traffic
  • end up killing the app after a year.

What has changed today is that for once, I wish to develop and launch an app which is solidly built and evolve it until it finds its audience. And that will take more than 3 months.

The long game

The plan is pretty simple. It fits both my professional identity as an academic who conducts and publishes research projects, and my evenings and week-ends interest in coding and developing apps, all while keeping a work / life balance. The idea is:

  1. develop an app that is super simple but done right: a one-page website where features can be easily added as independent modules, in the form of new pages or functions.
  2. add to it the features and functions that are laying in my drawers from my past research projects.
  3. develop new academic research projects, publish them.
  4. for each research project, publish an implementation of the method on the app, offered as a free service to whoever has an interest in it.

Take as much time as necessary for steps 1 and 2, then take 20 years or more to loop and repeat through steps 3 and 4, as many time as I want.

What I hope to achieve is a sense of aggregating and building up a coherent set of research outputs. The cumulative effect, I hope, will stimulate an interest in the whole, rather than just the parts. My research will be made available to a wide audience, which is another desirable outcome. It will live on a web platform that will grow and improve as time passes. It already gives me a feeling I appreciate a lot: looking far ahead, as when starting a long climb. Small steps, no rush, no timer. Playing the long game.

The app is Nocode functions. Check it now, and again in 5, 10 and 20 years :-)

 Date: July 8, 2021

⏪ Hello world

Pointwise mutual information and tf idf, when to use them ⏩