.png)
AI Hype Means Nothing If Your Data Definitions Are Still a Mess
- Christian Steinert
- Nov 18
- 5 min read
Why most healthcare companies need data management, not another AI tool
Everyone’s talking about AI agents and LLMs while most healthcare companies are still copying numbers from seven different places into Excel and calling it “the source of truth.”
This piece comes from the heart. It sits deep in my bones because it’s ultimately our offer at Steinert Analytics.
Since October, tech events have been a heavy focus for me. I’ve written about this in previous issues (specifically the one highlighting Kickstart Ohio 2025—where I met Steve Wozniak and rallied with other entrepreneurs).
We all know the drill—AI is the talk of the town. It seems like no matter where you go, AI is the most popular topic in existence. The United States labor market hasn’t seen this kind of disruption before.
As enterprises invest millions to billions of dollars into AI, naive professionals forget a few critical things as we cling to any job security we think is left.
What dominates the news headlines are market-leading enterprises like Salesforce, Amazon, Target, and Google. We automatically assume AI is taking over all areas of corporate life, but we forget where the bulk of companies are in their AI Maturity Lifecycle.
Most Companies Are Still Running Like It’s Elementary School
Leave it to my consulting experience to give me a first-hand look into how the majority of small to mid-market healthcare companies are running in “the Age of AI.”
I’ve written extensively about this on LinkedIn already.
The majority of these companies are still piecing together metrics from 7 different places and hard-coding the values into an Excel workbook. This gets distributed company-wide and is considered the “source of truth.”
There is barely any automation happening. And even more concerning, the information management discussions that should have happened years ago are only just beginning.
When I say “information management” discussions, I’m talking about defining what a data point means. This could be something as simple as how a healthcare company refers to a customer (aka customer vs. patient), or something a little more nuanced—calculating net revenue.
Do we include or exclude write-offs as part of our net revenue calculation? Which adjustments do we take off our Charge Amount to get our true “net”?
For those not in the data management space, you’d be SHOCKED at the number of companies that don’t truly define these metrics out of the gate.
I’m speaking from 8 years of experience in the space, working with every company profile from SMB to F200 enterprises.
Operations calculates it one way.
Finance another.
Marketing yet another.
All similar, but slight variations that make anywhere from a couple thousand to millions of dollars in difference.
This matters a shit load when you have a group of investors and a CEO in a board room talking EBITDA and valuations.
Each Metric Needs Alignment and Agreement
It might sound easy to establish these ground rules, but let me tell you—it isn’t.
Why?
Humans have opinions, perspectives, and egos that push back on rules (aka data management) like an African Bull Elephant bulldozing a tree.
Meetings I’ve Experienced
Let’s set a short meeting to get alignment on how we define Cash Goal. Do we look at it from the perspective of all cash collected in the last 2 months (60 days)? Or the last 7 days?
Sometimes it takes a company until the end of each month to collect on accounts receivable. There can be a mad dash at the end of a month to get invoices paid by a payor or private payor.
Therefore, if we’re measuring the efficiency of a company to collect physical cash from its sales, one perspective is that it might not be the most accurate view to consider monthly timeframes when the end of the month is a push for paid invoices anyways. Even if that’s industry standard.
You should consider cash collected each week to get a better picture for just how fast a company collects cash in some perspectives.
Both make sense, but differences in perspective, opinions, and egos make this seemingly straightforward conversation last over an hour.
A VP of Finance holds steady on the industry standard metric. A CEO holds firm on the metric that will actually indicate long-term cash flow efficiency.
See how this can get extremely complicated now?
That’s what we data management professionals facilitate and guide a business in doing. And it’s a whole lot more than just building the correct SQL code.
How Steinert Analytics Alleviates This Gap
I’m not going to pitch our entire offer. But I will speak from our experiences and the benefits our small to mid-sized healthcare clients have.
We kick off an engagement by documenting your entire data ecosystem. This typically isn’t leveraging some expensive and fancy data governance tool. In fact, the data foundation sprint entails populating a robust data catalogue in Google Sheets.
It’s cost-effective and puts the onus on us consultants to genuinely populate and update this living and breathing asset for your organization. We give it the attention and care it deserves, without breaking your bank.
In doing that, it surfaces lots of legacy logic (if brownfield). What a company calls Net Revenue we may discover that it’s defined in 3 different ways across 5 reports.
We document that and save it for the day we inevitably begin work as your healthcare organization’s fractional data engineering team of choice.
These steps give us the awareness we need to ask the right questions. Where we see differences, we also see opportunities to deploy data management exercises with your decision makers.
This involves calling out these inconsistencies as questions and orchestrating meetings that address discrepancies in definition. Ultimately, as painful and productive as these conversations are, we get your organization on a track of unparalleled trust in your data.
Whatever we build mirrors the people, processes, and technologies that support your healthcare organization’s long-term growth.
Correct logic.
Consistent definition.
Trusted data.
Robust insights and AI models that drive profit and generate revenue.
The Bottom Line
Without the unsexy and foundational work data management & engineering professionals do, all the hype around AI will fall short for your organization.
And unfortunately, that’s where the majority of companies are at right now.
Start with the unsexy stuff. Get your definitions aligned. Build trust in your data. Then—and only then—will AI actually move the needle for you.
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 email course here.
Comments