Do More with Less: The New Operating Model for Data Professionals
What worked before works no more
In the age of AI, data teams are expected to do more with less.
It's the reality for my analytics team, which is now a full-stack data team. In addition to creating metrics, conducting analyses, and building dashboards, we own the majority of our data lake: upstream data models, development in dbt, data ingestion, etc.
Things we used to rely on engineers for we do ourselves.
You see it too.
Organizations are flattening, the barrier to entry for technical skills is lowering, and leadership teams are expecting you to branch out into other areas.
The worst thing we can do is sit back and watch things shift under our feet.
The best thing we can do is develop a plan to learn to operate in the new world.
The "Do More with Less" Model
To do more, we must do things we've never done before.
We're expected to learn new skills, use new tools, and own work other teams used to do. This growth is easier to do this when it's forced upon you and harder if you need to do it proactively. Either way, it's time to lean into the discomfort of learning something new.
There are three steps we can take to do more with less:
- Make time to experiment with AI
- Use AI to automate parts of our job
- Identify roles that are adjacent to ours and learn them
Each step you take will increase your confidence operating in the changing landscape.
Make time to experiment with AI
This is a non-negotiable.
You can't know how to get value from AI if you don't experiment with it. The paradox is the time you spend experimenting is time you're not doing your core job responsibilities, and that feels wrong. Not immorally wrong, just misaligned with the way we've operated for years.
The mindset shift is to realize that using AI is part of your job responsibilities now, whether or not it's been stated explicitly. You must operate under the assumption that your manager expects you to use AI because it's definitely what the job market wants. It's critical to set time aside to learn AI.
- Block time on your calendar
- Take an online course
- Download Claude Code on your computer
- Build a Claude Skill from scratch
Once you experiment, you'll naturally think of ways to incorporate AI into your day-to-day.
Use AI to automate parts of your job
When we're expected to do more of one thing, we have to do less of something else.
There are two ways to do less:
- Stop doing work altogether
- Automate work
If we stop doing something, we aren't automating it. We are deciding to retire a type of work, project, or task. There are a small number of things that fit this criteria, so I'm more interested in automation.
Automation has multiple benefits, especially if we use AI to assist in that automation:
- We learn how to use AI to do real work
- We open our time to do other work
- We operate faster because tasks are automated
- We keep our commitments by keeping recurring tasks around (even if they're automated)
Automation is the way we create the space to "do more".
We have the same hours in a day, but our time is spent differently.
Identify adjacent roles and learn them
After we've learned AI and automated parts of our job, we can deploy our newfound time to doing new things.
Take my analytics team as an example. We were doing some data engineering and analytics engineering work, but now it's become a real responsibility. It was an adjacent role of ours and we absorbed it.
We all need to be proactive in learning and working within an adjacent area.
Analysts sit next to data engineers. Data engineers sit next to software engineers. Operations people sit next to analysts. If you work closely with a group, branch into their area.
As you stack your skills, you'll be valuable as the industry changes.
It's not pretty and there is no perfect path.
The only path is proactive action.
Use AI to grow, learn, and build skills that build a resilient data career.