.png)
“Do You Get Value From Your Data Team?” — “Not Really.”
- Christian Steinert
- Jul 7
- 6 min read
Where data teams stand in the Age of AI, and who actually survives.
Almost a decade in data. I’ve worked at corporate enterprises, mid-sized companies and start-ups and seen a fairly diverse set of data team environments as a consultant. What purpose do data teams serve?
I’ve been having plenty of conversations on this and experiencing a lot of interesting shifts in the market for the future of data teams in organizations.
Here’s how I’d define data teams in each environment based on my experience:
In the enterprise, it’s an IT function that owns the implementation of data integration, consolidation of sources and reporting.
In a mid-sized company, more of the same. It’s largely an IT function. Reporting, reporting, reporting. As Sebastian Hewing refers to it - a Dashboard Factory.
A start-up is where data teams create the most value. You’re working under either the Finance or Product department. The organization views you as more of a strategic enabler than just an IT-focused report creator and owner.
Based on my experiences, I’m laying out observations for where data teams stand in the Age of AI. Do they have a place inside a company or not?
The $75M Question Nobody Wants to Answer
On the surface, the need for a Data Team is a no-brainer to technical avatars of mid-sized to large organizations. Data is being generated from all angles, multiple source systems (EHR, CRM, Accounting/Budgeting), siloed Excel spreadsheets and various workflow automations (lead capturing into a CRM for example).
However, in the enterprise and mid-sized businesses, there is not enough strategic thinking. The data team ends up pulling in all of this data and plopping it into a data lake/lakehouse. As we’ve called it for many years now, it turns into a data swamp.
Data and analytics engineers build what they’re told to. KPIs without business definition, data models that have no applicability to a stakeholder’s actual pain, and a whole lot of guessing and hasty assumptions to boot.
In both my FTE and consulting life, this breakdown happens at the bridge between the business and the data team. You typically have a product manager interviewing the business for requirements. Somewhere along the way, the translation for value gets lost.
Either an inexperienced product manager who struggles to ask the right questions to identify a problem worth solving. A business analyst that sucks at writing acceptance criteria. Or engineers that only care about tools and build without understanding their business’s value chain.
Beyond strategic misalignment, I think a hard truth is that data teams just move too slow for business in the Age of AI (they’ve moved too slow for business even before the AI boom). I understand the push to build things “the right way”. I understand the need for best practices in data modeling, data visualization design and database optimization. However, is something truly “right” if it delays time to value and eliminates the chance for buy-in from the business? I talk more about this in one of my articles that Bill Inmon called “a masterpiece”.
Due to all these factors, I had a conversation with a founder of a 400 person, $75M revenue mid-sized scale up at a healthcare conference a few weeks ago. We were at a bar at 11:30 PM (funny enough, neither of us were drinking alcohol). After chatting about how he led his engineering team to beef up their AI skills, I asked him about his data team point blank - do you get value from them? His answer was “not really”. He then went into how complex his organization’s processes are. The failure to account for the tribal knowledge of workflows and how data is generated internally is the root cause of his discontent. I’d have to imagine the lack of business translation by the data team and slow time to value adds to this friction he feels (Big surprise! Haha).
Where Data Teams Earn Their Keep (And Where They’re Dead Weight)
On the flip side of the value gap, I’ve seen and learned about data teams that work well in the Enterprise for two main reasons.
Leadership at the top prioritizes data heavily
If you don’t have buy-in from the top down, just forget it. You’re always going to be fighting for requirements, time and resources to build things that actually help a business grow revenue, increase profit and enhance a product/service.
Governance and Master Data Management
If leadership prioritizes it, data governance can be extremely effective for defining KPIs and detail fields. Every definition and change runs through the data team, flowing through an approval process to ensure the data reporting source is always updated and representing the agreed upon definitions. This is extremely rare and hard to find a culture like this.
In the start-up world, I don’t think a data team is entirely necessary. Many startups scale on Google Sheets and a CRM. It’s not enough digital source systems to justify the overhead a data team carries and builds.
Many of the issues faced in these startups/smaller companies can be solved with a heavy focus on Revenue Operations and other core business processes. So many of them do not have their sales process defined (in my experience). Setting up their CRM and getting data generated correctly is the first step. A data infrastructure/team is far too out of scope at this maturity.
Any solutions you build at this stage can most likely exist in a Google Sheet or Excel.
You DO NOT need a full blown data infrastructure with governance yet.
I think data professionals need to put more energy into functions like RevOps. I’ve been chatting extensively about this with Ergest Xheblati. So extensive, that it’s re-shaping the offer at Steinert Analytics. It’s one of the cleanest ways to establish value as a data professional while satisfying their need for quick turnarounds. Build that trust, and you’ve got a seat at the table.
If a start-up is extremely data centric (telehealth, for example), the data team could technically be part of the product or finance team. This was, in fact, the structure at the first healthcare company I ever worked for as a lead data analyst.
Their goal is to build data assets that help with product-level automations. It’s about using business logic and criteria defined in data models to produce outputs that trigger an action for revenue or a better customer experience. (ie. automating billing codes for healthcare claims or sending patient documents to the correct providers, totally automated).
I don’t believe it takes the overhead of a data team to build these things. Either roll one or two data platform engineers into your engineering team, or hire a highly specialized consultant/boutique consultancy to own these data platform builds and the reporting layer. You’ll stay leaner, move faster and operate with more precision to the business needs because of their out of the box expertise and ability to ask the right questions to the business directly, without all the noise of a team structure and hierarchy dynamics created with a data team.
The Verdict: Leaner, Faster, or Gone
So all in all, where do I think data teams are headed?
Let’s start with the biggest shift: viewing data development as an engineering function needs to die. Having middlemen between the business and engineering is over. The best data professionals are both technical and great communicators. Leveraging AI helps with the communication side as well - prepping you on questions to ask and becoming more familiar with a business’s value chain quickly.
From there, the verdict by company size is simple.
Enterprise and mid-sized companies? Data teams stay relevant. Governance and master data management aren’t going away - not with the vast amount of source systems these organizations operate with.
Startups? Revenue ops workflows paired with quick and dirty data solutions. That’s it.
Data-centric startups (like a telehealth SaaS)? Data automations tied to value producing workflows - owned by a couple versatile engineers on your platform team or a highly specialized consultant.
Leaner. Faster. Without all the team structure fat.
Christian Steinert is the founder of Steinert Analytics, helping healthcare organizations turn data into actionable insights. Subscribe to Rooftop Insights for weekly perspectives on analytics and business intelligence in these industries.
Also - check out our free Healthcare Analytics Playbook eBook course here.
Comments