It seems like data/analytics has much less clarity around roles than most other common functions at tech companies. Do you think this will ever change?
Yes and no.
On one hand, niches are deepening. Cutting edge applications of data science like deep learning are getting more advanced, and have already led to some clear specialization among data scientists. Even within deep learning, there are computer vision specialists whose roles are extremely well-defined.
On the other hand, data literacy and skills are becoming much more common, especially among people whose primary job function is not analysis. We all already do analysis every day. Anytime you look for patterns in previous experiences or make a tradeoff between one option and another, you're analyzing. But the standard of what's expected as base-level analytical prowess is rising. 10 years ago, most of us were not on the hook to collect or interpret data. Today, we're at least expected to be able to get basic info from an application and interpret it to inform our decision making.
For many of Mode's customers, the number of people writing queries is often much higher than the actual number of full-time analysts at their companies. I've seen large companies where more than half the employees are writing SQL to get information. Facebook was like that was 8 years ago when I was there, and now there are a lot more companies like them. More and more folks outside of analytics and data science teams are making the jump from basic data literacy to some level of proficiency.
This increase in data proficiency across people of all backgrounds is making the Analyst role a lot harder to define, and possibly even more ambiguous than it is today. As this happens, it's possible that we will see some additional niches crop up, or that the notion of a general analyst evolves into something different. I think we're still many years from this crystallizing.
It'll mostly depend on the skills of other folks in the organization. The more data-savvy the rest of the company is, the more full-time data team members can focus on specific problems or skill sets.
For what it's worth, I consider it a very important part of Mode's mission to increase the average person's data proficiency. The ambiguity of these roles isn't necessarily a bad thing.