Analytics Engineers: It’s Time To Pick A Side

Analytics engineering started with a promise: bridge the gap between raw data and business insight. Cloud warehousing and open-source tools like dbt made this promise possible. But somewhere along the way we got stuck, trapped in SQL pipelines and process decks while decisions happened elsewhere.

Why? As always, things changed. The stack grew more complex, costs piled up, and the role became too narrow—focused solely on writing SQL and warehouse development.

I’ve seen this play out across dozens of companies. Teams are doing solid technical work and all the “right” things on paper, but they’re stuck in the middle, nowhere near the decisions that matter. And when that happens, the value is lost.

The best analytics engineers know how to make an impact. They don't act like middlemen; instead they take ownership of the entire lifecycle of data. It's not just ingestion, it's about giving the business insight and helping them drive towards better decisions.

Time to make a choice—you can’t stay in the middle

While there are many definitions of this role based on any given company, analytics engineers generally sit between upstream data engineering and downstream decision-making. That middle ground used to feel strategic but now it's becoming more of a bottleneck.

What you can’t do is sit passively in the middle, owning transformations but disconnected from both context and impact. In a few years, that role won't even exist.

Analytics engineers need to make a choice.

You can lean left—get deeper into ingestion, pipelines, and architecture. Understand where your data comes from and how it’s shaped. Or lean right—embed with the business. Sit in their meetings. Learn their goals. Translate messy questions into models that drive outcomes. 

What you can’t do is sit passively in the middle, owning transformations but disconnected from both context and impact. In a few years, that role won't even exist.

Be more than a builder

Some believe analytics engineering should run like a product org: roadmaps, planning, feature releases. That structure helps. But I’ve been in places where we followed that model too rigidly and blocked others from doing their jobs as a result.

And in today’s environment, speed matters. You don’t need to be perfect. You need to be 85–90% correct. If your team is bogged down in process and multiple rounds of code review for trivial changes, you’ll never move fast enough to make a difference.

The best teams strike a balance between strategy, service, and speed. You plan like a product team but stay agile because you never know when things will catch fire. 

On my team, we run an on-call rotation for ad-hoc requests. We don’t ignore asks just because they weren’t on the sprint board. Remember, our goal is to drive the business forward. Ad-hoc requests have impact and that matters just as much as supporting our key initiatives. Our responsiveness builds trust, and ask anyone in data, they’ll tell you trust is the metric that matters most. 

The best teams strike a balance between strategy, service, and speed. You plan like a product team but stay agile because you never know when things will catch fire. 

And trust is earned, and it starts with showing up. If you’re not included in stakeholder meetings, not consulted on decisions, not seen as essential—you’ve already lost. The relationship is the work.

Good data teams don’t wait to be included. They proactively make things clearer, faster, better. They make themselves part of how the business runs.

Don’t get stuck being a goalie playing defense

Legacy BI has made this kind of partnership impossible. They force data teams into defense mode—fielding requests, managing expectations, shielding their team from a flurry of questions.

But if people are asking questions, that’s not a problem. That means they’re curious and getting value from the data. They’re getting answers to their questions. Your job isn’t to gatekeep. It’s to keep that curiosity alive and make it productive. That means enabling self-service in a way that’s fast, flexible, and governed.

But if people are asking questions, that’s not a problem. That means they’re curious and getting value from the data. They’re getting answers to their questions.

The best tools let you stay close to the business. They let you move fast when someone needs an answer and still build systems that scale. If your platform forces you to choose between code-only or drag-and-drop training wheels, you’ll always be stuck on defense.

And when that happens, you’ve only got two options: slow everything down with process or be left behind while decisions happen without you.

What good actually looks like

Done right, analytics engineering becomes a key lever for the business:

• You operate strategically, blending powerful cloud platforms with intuitive analytics tools to deliver governed self-service.
  
  
• You build interactive data apps that go beyond static reports, automating processes that drive collaboration and action.
  
  
• You work side by side with the business to understand needs and build workflows that unlock new possibilities.
  
  
• You use AI—whether automating analysis or adding natural language to dashboards, it’s now table stakes.

This framework makes it easy for our team to support the business and make an impact. We’ve got sales leaders forecasting in Sigma, financial analysts making budget adjustments, and execs modeling top-line metrics like ARR, GRR, and NDR—all directly in the product. They’re not waiting on a report. They’re interacting with live data, asking better questions, and making faster calls.

The most important thing you’ll read today

You don’t earn trust with tools and there’s no guarantee that a six-figure spend on warehousing will solve your problems. Instead, you need to start with discovery.

Talk to your stakeholders, your CEO, your sales leaders, and operation heads. Ask them where they’re getting stuck, what’s too confusing, what gets exported to Excel just to become outdated?

That’s your roadmap. That’s how you’ll make an impact. That’s the future of your job. 

‍