Is the keyboard mightier than the mouse?
When it comes to designing tools and workflows and generally thinking about how biotech data teams should work, it’s often useful to split up users into two buckets that I call keyboard users and mouse users. Keyboard users are the ones who like to write code. They store data in CSVs, parquet files and databases, and share analysis results through Github links and Jupyter notebooks. Mouse users are the ones who prefer point-and-click tools with graphical user interfaces. They store their data in spreadsheets and BI tools, and share their analysis through slide decks and spreadsheets, sent as email attachments.
Biotech bench scientists tend to be mouse users (though there are always exceptions, as well as folks in between the two buckets) while data teams tend to be keyboard users. So most biotechs face the challenge of finding ways for these two types of users, with very different preferences and types of tools, to work together, to share data and results. Anyone who has tried to share a Jupyter notebook with a colleague who doesn’t have a Github account knows this.
It’s generally easier to build custom tools for keyboard users because the kinds of interfaces that they want - CLIs and Python libraries and REST APIs - require less overhead than the point-and-click UIs that mouse users need. So internally-built tools tend to support keyboard users over mouse users.
Commercial software, on the other hand, tends to be built for mouse users. This is partially because it’s easier to sell a UI-based product to mouse users than to sell a library or an API to keyboard users. But it’s also because the developers recognize that they need to support mouse users and it’s easier to justify the cost of building a UI as a commercial product vs an internal tool.
Of course, these commercial options are mostly built BY keyboard users, so they tend to build in things like REST APIs that keyboard users will want. The problem is that they’re less flexible, by their nature, than open source libraries and home-built tools.
So it’s very hard for these tools to make the immediate work easier for keyboard users compared to these other options. And frankly, it’s almost as hard to make the immediate work easier for mouse users compared to just using a spreadsheet. Instead, what they have the opportunity to do is to become the common tool - the glue - that allows both kinds of users to work together. And that’s what we should look for when we evaluate them.
Scaling Biotech is brought to you by Merelogic - an independent consulting firm that helps biotech startups design processes, conventions and digital tools to break through the tech problems that come between them and the science.