"Bringing data” simply isn’t good enough when setting benefits strategy 

“In God we trust; all others bring data” – W. Edwards Deming

That quote is a favorite of data nerds from all walks of business. It has the benefit of being memorable, pithy, and almost accurate. I say “almost” because just “bringing data” isn’t good enough anymore. For data to be useful, it needs to be cleaned and normalized, it needs to be integrated and aggregated, it needs to be enhanced and scaled, but most importantly, it needs to be understood.

Consider the complexity of your benefits ecosystem today. You are managing one or more health plans. You may have carved out pharmacy benefits to a PBM. You may have point solutions addressing women’s health or specific conditions like diabetes, and COEs to address complex conditions. You may offer an advocacy program to help your employes navigate it all seamlessly.

All of those programs generate data. Financial data, clinical data, outcomes data, behavioral health data – and all of it is necessary for us to understand if all those solutions are working as we intended. Where there are opportunities for adjustments and course corrections, we can identify them early.

All of this may seem obvious – why wouldn’t every decision concerning an annual operating expense in the tens or hundreds of millions of dollars be firmly based on data? The simple answer is that in our space, data is hard. There is limited standardization of data layouts, and limited transparency between data suppliers and their customers. The good news is that, as challenging as it may be, the industry has made significant progress, and you can take advantage of these advances.

Here are a few best practices in using data for measurement and management:

Just get started. If you haven’t started down the path of a formal measurement strategy, don’t be overwhelmed! You don’t have to go from zero to a fully integrated data warehouse with dozens of feeds overnight. Inventory the data you have, understand the gaps, identify what’s usable, and then, simply, use it. One thing about data, it always creates the appetite for more data.

Identify your expectations. Every strategic intervention we make has an intended outcome. Whether we’re changing a deductible, adding a care management program, or changing our health plan administrator, there’s always an expected result. It is essential that we understand these goals before we launch a program. Are we trying to improve health? Impact productivity? Save money? All of these are valid goals, but if we do not articulate them – and the specific data we’re going to use to measure them – we won’t know if these strategies are successful. It is also essential to recognize the specific timing associated with our expectations. While we may expect to see cost savings fairly quickly, health improvement may take a year or longer to show up in the data. With clearly defined leading, concurrent, and lagging success indicators, we can see trouble coming, and pivot to address these issues before they undermine our strategy.

Consider data warehouse solutions. The value proposition for independent data warehouses has never been easier to define. The cost of an effective data warehouse may be significantly lower than you think, even if it has only been a short time since you explored your options. A single, objective platform that allows you to integrate your own data; enhance it with clinical methodologies, risk scores, and social determinants of health; and efficiently generate flexible reports will quickly become an indispensable tool.

Be ready for the future. AI, machine learning, and large language models like ChatGPT all depend on data to deliver on their promise of providing data-driven insights to our most complex challenges. The pace of evolution of these tools has made the case for organizing your data even more compelling. You will continue to see your consulting partners, health plan providers, and point solution vendors wanting to use data to enhance their solution sets.

While there’s new urgency in the way we talk about data, the truth is we’ve known how important data is for decades. After all, Deming did his work on quality improvement in the 1950s. It’s a fair criticism that the benefits function, and perhaps HR more broadly, has been slower to adopt disciplined measurement practices than other business functions. We can no longer watch from the sidelines. Get into your data, see if you need more of it, figure out what it’s telling you, and you’ll be on your way.

This post is one in a series of resolutions to guide benefit strategy.