4 Signs Your Company Actually Needs a Data Warehouse

The readiness checklist that separates strategic investments from expensive mistakes

Data consulting leads to interesting experiences. Those often involve a range of companies with large variations in tech and data maturity. If there’s one thing I’m learning, it’s knowing how to spot if a company even needs your services in the first place.

At Steinert Analytics, we specialize in building data infrastructure and governance programs for mid-market and scale-up healthcare orgs to drive data trust, literacy, and capability that increases profit and benefits patient outcomes.

As data leaders, it’s easy to fall into the trap thinking everyone needs robust data infrastructure yesterday. With the demand placed on data teams to deliver reporting and insights, we see and understand the value a data warehouse drives daily.

However, not all companies will immediately benefit from such an implementation.

The most critical question we ask: Is the prospect we’re talking to even ready for a data warehouse & strategy to begin with?

You’d be surprised at the lack of maturity of some healthcare organizations. Filled with manual processes and hard-coded spreadsheets, a data warehouse out of the gate does not always make sense.

In this issue, I want to cover a few signs that a company will actually benefit from a data warehouse.

4 Signs a Company is Ready for Data Warehousing

Sign 1: There’s More Than One Source System

Does the company have a system of record (i.e., CRM, billing & accounting, EHR)? If it’s only one, most of the value in data and reporting can be found in that system’s native analytics capabilities.

Obviously, it depends on the vendor, but I’d always encourage an immature company like this to squeeze as much value out of their native analytics solution tied to their source system.

The con of this approach? If multiple users are exporting data from the source and manipulating it in their own way, metric definitions can become twisted. There is no governance.

However, typically companies that only have one source system are small (in my experience). The juice isn’t worth the squeeze to try and implement an enterprise-grade data architecture and governance strategy. It costs far too much and isn’t completely necessary.

If there is more than one source system (i.e., having a CRM and accounting software), now we’re getting somewhere. Often, companies need data pulled in from both the CRM and accounting software to accurately track revenue.

The CRM could have a varying definition of revenue compared to the accounting software. This is where modeling and conforming the data together in a single combined table is super useful.

Furthermore, if you want to understand the full picture of a customer’s journey to land on an accurate Customer Lifetime Value metric, it’s necessary to track when a customer was acquired via a given marketing campaign (CRM data) and the actual amount of revenue from that customer (financial system data).

Sign 2: The Data Infrastructure Can’t Keep Up With the Company’s Scale

Situations like this commonly arise when companies have a legacy on-prem system. Typically a SQL Server database powering their “data lake” (more like a data swamp, haha!).

It might be anywhere from 250GB to 1TB worth of data. In healthcare, this typically contains EHR, call center, and financial general ledger data (think NetSuite). There might be 50 to 100 stored procedures that kick off jobs to load report tables that are a hack show of a data model at best. Scattered queries, metric logics, and inconsistent join patterns make it hard to determine what is correct.

Metric sprawl and no data governance at its finest.

Loading these jobs—hundreds of millions of rows—puts strain on the database. Often jobs get stuck and fail.

One client we saw had to implement a SAS backplane in order to increase memory enough to keep limping along their on-prem database while we built their data warehouse in Microsoft Fabric.

Sign 3: Your Operational Workflows Are Digitally Established and You Want to Realize Deeper Insights From Them

Don’t put the cart before the horse. I spent an entire year trying to force data infrastructure into the roofing and home services industry without truly understanding their problems (hence my pivot into healthcare).

The early lesson learned there was exactly this point. Companies need to be further along the AI maturity lifecycle before they need data infrastructure. If you slap a data solution on top of a broken digital system, it’ll produce garbage. “Garbage in, garbage out” as they always say.

This comes down to identifying processes and workflows using their source systems. What’s broken? What’s inconsistent?

For example: How does a sales rep correctly add a lead into the CRM? Is this done consistently every time? Is every field needed to understand a lead, opportunity, or customer accurate and complete, or is only part of the picture available?

This all drastically impacts the quality of the data we have to work with in a data solution. If that quality is poor, DO NOT waste time building a data warehouse yet. They aren’t ready.

I’ve seen similar situations in healthcare. Broken processes that lead to poor data quality in the CRM and financial software.

Focus on fixing those problems first. Worry about the data infrastructure once those are in place.

A great exercise to get stakeholder process alignment is a Business Data Model (BDM). I’ve done this for companies in all types of industries. There’s a write-up about the value a BDM provided a client of ours in the roofing industry last year (link).

Sign 4: Your Company Is Made Up of Multiple Departments That Function Separately From Each Other

I’m pulling this one from Bill Inmon’s bucket of knowledge. I vividly remember him mentioning this on The Joe Reis Show podcast a while back (link). It stuck with me.

Point 4 is my own synthesized statement, but Bill dove into what a data warehouse really is (see a paraphrase below):

It’s a repository of corporate or enterprise data. Marketing, finance, accounting, and management departments that need to see a plethora of data. This data lives in separate applications and needs to integrate with each other.

My thought is: Confirm signs 1 and 4 and there’s a GREAT chance a data warehouse is needed at your company. Notice Bill’s focus on corporate/enterprise. Maybe not as relevant for smaller companies, but extremely valuable in data-heavy industries and companies with multiple source systems all being used by different departments.

Healthcare/healthtech is FULL of use cases for a data warehouse. The sheer amount of source systems generating data is overwhelming (at least for me). Niche EHRs, financial systems, telehealth usage, clinical data, insurance/claims. The list goes on. Pair that with the vast amount of departments doing different workflows with these systems, the picture starts painting itself for company-wide data integration and conformity with a warehouse.

The Bottom Line

There you have it. 4 signs a company is ready for data warehousing (with a focus on healthcare companies). I hope this helps paint a clearer picture for you.

I’ve been in so many situations where people wanted to build a data warehouse to “see the data better.” However, they failed to consider these critical aspects that signal it’ll actually be valuable.

Data warehousing is valuable, but it depends on the company’s maturity, technical sophistication, and questions they’re trying to answer from their data.

Always keep these in mind as you enter into a company looking to enable them with data & insights.

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.

Feel free to book a call with us here or reach out to Christian on LinkedIn. Thank you!

Also - check out our free Healthcare Analytics Playbook eBook here.