One change at a time

In car racing, drivers are taught not to hit the brakes and turn the wheel at the same time. The tires don’t have enough traction to make major changes to both direction and speed at once.

Integrating a data team into a biotech organization requires a similar rule for changing the technology and the processes around it. New software for collecting sample metadata require new lab processes to collect it at the right point in the experiment. New analysis tools need to be incorporated into bench scientists’ post-experiment workflows. New predictive models require new ways of designing and evaluating experiments.

Any change takes time and energy to get people on board and moving in the same direction. But changes to technology require a different kind of time and energy than changes to process. One involves mostly design and coding. The other involves mostly talking to people. It’s possible to do both at once, but switching between the two modes can be quite difficult.

Often, there are ways to handle them in separate steps. Start with a mostly manual prototype to shift the process, before you build the automation. Build a version of the tool that fits into existing workflows before you shift the process to use it more effectively. By focusing each change on either technology or process, you can maintain enough traction to stay on the track.