I’ll Take “Data Strategy” for $1,000 Please, Alex!

On Jeopardy – sorry to the new hosts, but Jeopardy will always be about Alex Trebek to me – knowing the right “questions” to answer is the key. The same is true for your data strategy. Over the last three weeks I have made the case that AI is only as good as the data you feed it, that humans need to stay in the driver’s seat when it comes to reviewing AI recommendations, and that the five layers of the Data Hierarchy Pyramid give you a clear picture of what a solid facility data foundation actually looks like.  

Today I want to wrap up the series with the most practical question of all: how do you actually build a data strategy? 

At Roth IAMS, we have been helping organizations develop what we call their “desired dataset” since the company started over a decade ago. The concept is simple. Before you decide what data to collect, you need to know what questions you are trying to answer. Your data strategy should flow directly from those questions, not the other way around. Too many organizations collect data because it is available, or because a vendor told them they should, and then wonder why they cannot do anything useful with it. 

So here are the key questions we recommend every facility and asset management organization ask themselves as they build or refresh their data strategy — with AI firmly in mind. 

1. What story are you trying to tell?

This is always the first question, and it is the one most organizations skip. Are you primarily trying to make the case for increased capital renewal funding? Are you trying to build a defensible multi-year capital plan? Are you trying to reduce reactive maintenance and build out a Preventative Maintenance program? Are you trying to meet sustainability and greenhouse gas reduction commitments? Are you trying to address accessibility compliance obligations? The data you need, and the order in which you should collect it, depends entirely on the answers to these questions and the narrative story arc you want to take to your stakeholders. If you try to build your data strategy without understanding the story or stories (if you have different stakeholder groups) you are most likely going to end up missing the boat. 

2. Where do we currently sit on the Data Hierarchy Pyramid?

Last week I introduced the five levels of the Data Hierarchy Pyramid: Equipment Inventory, Facility Condition Assessment data, Regulatory and Code data, Programmatic and Functional data, and Sustainability and Energy data. An honest assessment of where your organization currently sits on each level of the Pyramid is the essential starting point for any data strategy. For each layer, ask: do we have this data at all? Is it current? Is it consistent and defensible across our entire portfolio? Is it integrated with the other layers? The gaps you identify here become the roadmap for your data collection priorities. 

3. How confident are we in the data we already have?

This is the question most organizations are afraid to ask honestly. Existing data is only an asset if it is reliable. Legacy datasets, old FCAs, spreadsheet-based equipment lists, and maintenance records that were never properly coded, can create as many problems as they solve if they are fed uncritically into an AI system or a capital planning tool. Before you build on top of what you have, take the time to understand its limitations. How old is it? Was it collected consistently across the portfolio? Can you defend the methodology behind it? In some cases, starting fresh is more cost-effective than trying to rehabilitate data that cannot be trusted. 

4. What standard will we hold our data to?

Consistent and defensible is the standard we apply to all facility data at Roth IAMS, and it is the right target for any organization serious about AI readiness. But those words need to be translated into specifics. What level of FCA detail do you need, Uniformat II Level 2, Level 3, or Level 4? What equipment types are in scope for your EI&T program, and what data points does each piece of equipment require? What regulations/standards will your accessibility assessments follow? What energy metrics will you track and how will you normalize them across your portfolio? These decisions need to be made deliberately and documented clearly because once you set a standard, everyone collecting and maintaining data across your organization needs to work to it. 

5. Who owns the data and who is responsible for keeping it current?

This is where a lot of data strategies fall apart in practice. Data collection is often treated as a project, something you do once, hand off, and move on from. But good data is a living asset that requires on-going stewardship. For each layer of the pyramid, your data strategy needs to assign clear ownership. Who is responsible for keeping this dataset current, how often it need to be refreshed, and what triggers an update outside of a regular cycle? Without clear ownership and a refresh plan, even the best dataset will go stale, and stale data fed into an AI is just efficient bad advice,  which brings us right back to where we started this series. 

6. What is our integration plan?

As I noted last week, the real power of AI in facilities and asset management comes when your datasets are integrated, when your AI tools can look across all your data simultaneously and surface insights that no single dataset could reveal on its own. Your data strategy needs to have a point of view on how these datasets connect to each other, what software platforms will hold them, and how they will be kept synchronized as your portfolio changes over time. This is not a small undertaking, but it is the difference between using AI to answer individual questions and using AI to genuinely transform how your organization makes decisions. 

Building a strong data strategy is not a glamorous exercise. It takes honest self-assessment, clear priorities, organizational discipline and a long-term commitment to treating your data as the strategic asset it is. But organizations that do this work, that climb the Data Hierarchy Pyramid deliberately and with clear intent, are the ones that will be ready to unlock what AI actually has to offer. 

The AI era in facility and asset management is not coming. It is here. The “Haves” are going to be the organizations that meet it with a solid data foundation, engaged human judgment and a clear strategy for both. The “Have Nots” are going to find out the hard way that technology alone is not the answer.