How the Data role is evolving in 2026 (and beyond)
What you MUST know about skills, gaps, and opportunities
Over the past few weeks, I talked to dozens Data leaders and Data practitioners across different industries. Tech, entertainment, consulting, retail etc. I wanted to get a bird’s-eye view of how AI is changing the Data profession.
Some teams are running hard on AI, while others are still figuring it out. But no matter the level of adoption, one thing is consistent. The Data role is changing.
I’m sharing them because they point to where you (as a Data professional) might expect your career to evolve and how you might stay ahead.
Let’s get into it.
1. Soft skills starting to become the most important skills
We’ve always known that soft skills mattered in Data. Storytelling, framing the right questions, stakeholder influence, relationship management, communication etc. But more recently, these have moved to the top of the list of most important skills for a Data professional.
But why? AI has taken over the “traditional” technical skills, like SQL, Python, and coding in general. They’re (sometimes) even doing it better than we would.
One thing I would flag as the “new competitive edge” is what one Data leader called a “translation muscle.” Being able to take a complex technical concept and make it land with product, legal, engineering, or whoever you’re talking to. Data sits at the intersection of technical skills and the business, and this is what we are trained to do.
So if you’ve been putting off working on your communication skills because you’d rather just code, now is the time.
2. Constant learning is now table stakes
Once upon a time, you could get by on the skills you graduated with (plus a few years of on-the-job learning). Those days are gone. If you’re a Data professional in 2026, you need to be learning constantly.
And it’s on leaders to make space and time for this. Here’s what’s working across the orgs I talked to:
Protected calendar time (that’s mandatory for both the team and business partners to respect)
Peer-led monthly or weekly knowledge sharing sessions
Hackathons or learning sprints, where everyone on the team comes together, turns off distractions and spends time upskilling
Conference budgets tied to a knowledge-share back to the team
You might be thinking: “The technology is evolving so quickly that by the time you get good at one tool, the next one has already replaced it. So how do we keep up?”
You don’t.
“Wait what? Don’t keep up?” - what I imagine your reaction is
Well, don’t try to keep up with every tool. Learn the principles instead. Tools change every 6 months, but database fundamentals from 20 years ago still apply. Good data modeling, clean code, solid statistical thinking. This stuff doesn’t go out of style.
3. Governance is the biggest gap, and the biggest opportunity for YOU
Almost every company I talked to has rolled out AI tools to their teams. Copilot, ChatGPT, Cursor, Claude Code, internal assistants. But very few have shared standards, ownership, or guardrails around how these tools should be used.
Evals are a great example. Even at sophisticated orgs, evals are treated as an afterthought. And when they do exist, they’re usually superficial checks on tone and relevance, and not accuracy, sources, or tool-calling correctness.
So this is an opportunity for you as a Data professional to step up and lead. Governance conversations are (often only) happening at senior levels right now. Anyone who can hold their own across product, legal, and engineering is bringing a lot of measurable value to the org.
4. Data foundations (and the skills to build those) will decide who wins in the AI era
Here’s the thing about LLMs. They’re probabilistic. Which means they’re only as good as the data you feed them. If your data is fragmented, messy, or trash, your AI outputs will be fragmented, messy, and trash too.
One Data leader put it perfectly: “ask the same question five times, get five different answers.” That’s what happens when the data model can’t figure out the right join paths, tables, and filters.
On the flip side, teams with strong data foundations are going to be able to build production-level AI solutions with more consistent and predictable outputs.
My take on this.. Even if you’re not a Data Engineer, build up your Data Engineering skills. Do the boring difficult things, like writing data documentation, building semantic models, and auditing the data pipelines. It’s the highest-leverage work most Data teams can do right now.
I personally think this is what will separate the companies that will win this AI era from the teams that’ll get left behind.
5. Role boundaries are collapsing
The swim lanes between data scientist, ML engineer, AI engineer, analytics engineer, and data engineer are starting to blur. Data scientists are doing data engineering work. Data engineers are prototyping models. The clean org chart where each role is very clearly defined is not really how teams are operating anymore.
And it’s not just within Data. The boundaries between Data and non-Data roles are blurring too. I’ve seen PMs do their own analysis, designers pull their own metrics, and engineers build data pipelines.
So what this means for you.
If your niche is on the narrower side (like building dashboards or writing queries all day), it’s worth thinking about how you expand your skill set. Consider how you might build out more skills so that you become more of a Data generalist.
The Data professionals I see doing well right now are T-shaped. Wide enough to collaborate across disciplines, but with real depth in at least one area (that could be a technical specialty, or it could be deep domain expertise in your industry).
Resumes get skimmed for 6 seconds. If your experience doesn’t immediately match what the recruiter is looking for, you’re getting passed over.
So what can you do? Tailor your resume to every single job.
I know that sounds exhausting. But I’ve got it down to under 10 minutes, and it’s mostly automated.
Here’s my workflow using Airtable AI agents:
Create a Google Doc with ALL your past work experience (one-time setup)
Copy and paste the job listing URL into Airtable.
From there, the field agents take over:
↳ They extract the full job description and parse out the required skills
↳ A “Relevant Work Experience” agent reads your Google Doc and picks the 3-5 bullets that best match this specific role
↳ A “Resume Summary” agent writes a tailored highlights section for the jobAll you do is copy those columns into a fresh resume
Oh, plus a bonus! The same template can also generate a personalized website for each application if you really want to stand out.
Use the template I built with Airtable → https://www.airtable.com/lp/dawn
ICYMI (in case you missed it)
How to use AI to generate slide decks (that actually sound like you)
How to solve Data Science case study interviews
3 of my favorite AI prompting frameworks
Rating job search advice
Fun one! Kiss, marry, kill — data edition