Your Staff Is Drowning in Spreadsheets: A Better Way Exists.

"If you develop an integrated system but the culture of the organization is not around data engagement, those tools will sit on a shelf. Data engagement creates the culture. Integrated systems then scale and sustain it."

  -- Heidi Milch, Executive Director at CCNY Inc.

A conversation with Heidi Milch on why data integration starts long before technology.

Picture this: a case manager at a mid-sized human services agency starts her Monday morning by pulling a report from the case management system, copying numbers into a spreadsheet, then logging into two different grant portals to reconcile outcome data before her 9 a.m. supervision. By the time she sits down with her supervisor, she has already spent an hour doing work that has nothing to do with clients.

This is not a time management problem. It is a data infrastructure problem. And it is one of the most common drains on capacity that health and human service organizations face today.

We sat down with Heidi Milch, our executive director and data strategy expert, who has worked alongside health and human services nonprofits for over two decades, to talk about what data integration actually means, why so many organizations are not ready for it yet, and what the real cost of doing nothing looks like.

So What Is Data Integration, Really?

For most executive directors, the words "data integration" conjure up something expensive, complicated, and probably a project for next fiscal year. But the concept itself is more straightforward than the jargon suggests.

"An integrated data ecosystem takes all the sources of data an organization has and puts them in one place," says Milch. "So you can see the full picture, including clinical outcomes, HR data, payroll, and fiscal data, all exist together. That is when you can start asking the questions that actually matter, like what does it cost us to produce a positive outcome per client? Or what is the real cost of employee turnover versus the cost of retention?"

Right now, most organizations are making strategic decisions while looking at four or five separate slices of the same picture. The full picture is never in view at one time.

The Prerequisite Nobody Talks About

Here is where the conversation gets interesting, and honestly, where a lot of organizations get tripped up. Building a sophisticated integrated data system before your staff actually trusts and uses data is one of the most common and costly mistakes in this space.

"Data engagement has to come first," says Milch. "If the culture of the organization is not already around using data, asking questions about data, and feeling safe doing both, then a sophisticated integrated system will just sit collecting dust on a shelf."

Data engagement means frontline staff understand what the numbers they are entering actually mean. It means supervisors use data in coaching conversations rather than just pulling it out for grant audits. It means leadership makes decisions visibly and transparently from data, so staff can see the connection between their documentation and organizational direction.

And critically, it means staff feel psychologically safe questioning data. Milch shared an example she encountered recently, an organization where supervisors were presenting individual clinician data publicly in staff meetings without context, calling out performance in front of peers. The result was predictable. Staff stopped trusting the data. They stopped trusting leadership. And any hope of building a healthy data culture quietly evaporated.

"Trauma-informed use of data is not just a clinical concept," Milch says. "It applies to how you use data with staff too. If people are afraid of the data, they will never engage with it meaningfully. And then you have built an expensive infrastructure that nobody actually uses."

Data engagement, data literacy, and data governance are the foundation. Integration comes later, built on top of that trust.

What Disconnected Systems Actually Cost You

Beyond the time drain of manual reporting, data silos create a problem that often goes unexamined: the client who has to keep telling their own story over and over again.

"We get called all the time by organizations that do not know all the programs their own clients are enrolled in," says Milch. "You have someone in preventive services, also accessing the food pantry, also getting transportation assistance, also in substance abuse counseling, and none of those programs know what the others are doing. That client has to then re-explain their situation every single time."

David Monroe, who works alongside Milch and joined the conversation, added a layer to this: when data is siloed, that one client gets counted as five. Duplicated counts become unreliable, grant reporting becomes a problem and organizations lose the ability to understand what combination of services is actually producing outcomes.

"If you only look at one person in one program without seeing everything else they are receiving, you risk replicating that siloed service for the next client and wondering why you are not getting the same result," says Milch. "It takes a village. And without integrated data, you cannot see the village as a whole."

Monroe shared a story that brings this point home: one organization had their intake process updated separately by each program, with no communication across silos. The result was a massive intake that asked the same questions repeatedly. A client became quickly frustrated and angry with the process, desperately just wanting help. Everyone deserves to have quick access to programs, hopefully only having to share that story once.

When Integration Works: A Real Example

One of the clearest examples Milch and Monroe described involved a productivity dashboard that brought together clinician data from HR, scheduling, billing, and fiscal sources in one view. What emerged from that integrated picture was something no single system could have surfaced on its own.

One clinician was generating the most billable revenue on the team while working fewer hours than everyone else. The integrated data revealed why: this person had a dramatically lower no-show rate. Clients showed up. The organization did not need to hire a new clinician. They needed to understand what that person was doing differently with client engagement, and then help this clinician take on more clients, since the capacity was there.

None of that insight was available when HR, billing, and scheduling each lived in their own corner.

Another example involved integrating claims data, clinical records, and census data to identify which families were at highest risk for child removal, and then directing specialized programs and resources to the neighborhoods where that risk was concentrated. That is the kind of decision-making that is simply not possible when data lives in separate systems.

"I Simply Don't Have Time for Another System"

This is the most common objection, and it is worth taking seriously rather than dismissing. If integration means layering one more system on top of already overwhelmed staff, the concern is legitimate. But done well, integration does the opposite. It removes the time staff currently spend pulling data from multiple places, cleaning it manually, reconciling discrepancies, and building pivot tables in spreadsheets that need to be rebuilt from scratch every reporting cycle.

"Every time a human has to manually transform data, there is an opportunity for error," says Milch. "When it is automated and centralized, you know it is clean, you know it is consistent, and you know the definitions are standardized. The human in the loop shifts from cleaning data to actually analyzing it."

For smaller organizations without a dedicated data engineer or analyst, the investment in a well-designed integration service, with standard dashboards for regular use plus the ability to create custom queries from clean data, can save time at every level of the organization.

What Is Realistic Right Now?

For organizations with limited IT budgets and capacity, Milch is clear about where to begin: not with technology.

The road to data integration runs through four stages, in order:

Data mapping. Understanding what data you have, where it lives, and how it is being collected across every system in the organization.

Data governance. Establishing standard definitions. Who is responsible for data quality? How is each variable defined? Are two different programs measuring the same thing two different ways?

Data literacy. Do staff actually understand what the data means? Can they distinguish between outputs and outcomes? Do supervisors know how to use data in a way that builds trust rather than fear?

Data engagement. A culture where data is used regularly, operationally, and without fear, at every level from frontline staff to the executive suite.
"The cost associated with those first steps is time," says Milch. "Not technology. And I can guarantee that being intentional about data engagement will still reap rewards before you ever touch an integration platform. You will start doing quality improvement differently. You will start making decisions differently."

A Note on Choosing the Right Partner