One of the more emotional discussions with data scientists and analysts across the world tends to be around what tools they have access to. Perhaps it is because of the amount of time many tools take to learn, making it scary to contemplate moving to a new tool, as very few want to go through that learning process again.
Or even more significant, you used an amazing tool and were highly efficient and capable of producing spectacular results, then moved to a company that didn’t have a tool that was comparable. Certainly the loss of a world-class tool can feel like the loss of a superpower.
One of the key choices an organization must make when picking tools will be around use. Will the portfolio of data science tools include tools that are useful for the analyst or citizen data scientist, or only the data scientists? Will some tools be used by both, or will there be different tools for the various participants?
Consider Who Will Be Using the Solution
When I go to the mechanic or get a glance at a woodworker’s shop, I see a combination of tools in the toolbox. Many I recognize, know how to use, and feel comfortable with: the wrench, screwdriver, and a few others. And then there are some tools I couldn’t even name, like the adze.
Data Scientists and the Adze
If I tried to use the adze, I might not only be uncomfortable, I might significantly damage my leg. Who creates a tool that you swing towards your ankles with a sharp tip? Exactly how this tool has survived from the stone ages to today is beyond me, but it’s a beloved specialty piece for woodworkers and craftsmen.
The adze does, however, remind me of many data scientists who favor super-specialty tools that have little overlap with the tools beloved by citizen data scientists and analysts.
Many data scientists are programming in R and Python, with a smaller percentage using legacy solutions. These means are typically not accessible by most analysts, as the learning curve is fairly long with significantly low ease of use.
Analysts, Citizen Data Scientists, and the Wrench
On the other extreme, it’s common to see the analyst stuck performing a variety of jobs with only a wrench, aka the ubiquitous spreadsheet, which was not purpose-built to perform Big Data or advanced analytics. Since spreadsheets can’t do it all, analysts have to tap into the expertise of data scientists for certain tasks, which bogs down data scientists with projects that suck up their time and are often below their skill level.
Choose a Solution Built for a Variety of Skill Levels — the Swiss Army Knife
There’s a different choice that can be made, and that I have seen employed extremely successfully at several companies. Finding a data science tool that can be easily used by the entire community of users, from analyst to data scientist, allows new ways of working together.
This model creates a relationship where the data scientists become teachers and mentors and the tool becomes a platform for sharing and communicating. The level of understanding and transparency within models and solutions increases and the number of success stories increases rapidly.
Using a modern data analytics platform, where the data scientist can open a Jupyter notebook and write some Python or insert R code, while the analyst can drag-and-drop the most common functions without writing any code, speeds the development of the data scientist while giving amazing capabilities to the analyst.
At the end of the day, what you need is the ability for the platform to offer usage across the full spectrum of data workers.
Additionally, finding a modern analytics solution that garners amazing support from a community of users is even better and allows you the kind of 24/7 self-service support you may need.
Do you have a solution that can be leveraged across your enterprise? Do you have wrenches and adzes incapable of working together? Could both analysts and data scientists be doing more, together, with a Swiss Army knife?
Consider these five features to pick the right analytics solution for you:
- Ease of use
- Low investment
- Extensible across data roles
- Integration with data sources and technology
- Excellent support