Operational vs. Strategic AI Initiatives

Artificial intelligence in real estate is often discussed as a single category of technology, but in practice AI capabilities fall into two fundamentally different categories: operational capabilities and strategic capabilities.  Understanding this distinction is critical for organizations attempting to evaluate the feasibility of potential AI applications, allocate resources, and determine what kind of value will be created.  Many firms struggle with AI because they put these two categories of outcomes in the same bucket and try to evaluate them the same way.  They therefore misjudge expectations, required resources, and return potential.

Operational capabilities focus primarily on efficiency and automation within specific functions.  These applications typically target day-to-day execution and are designed to improve speed, accuracy, or cost efficiency.  Examples include report generation, data consolidation, document processing, and workflow automation.  These systems tend to be limited in scope, operating within a narrow functional boundary rather than influencing decision-making across the organization.  Because they are relatively contained, operational applications are usually easier to implement, less expensive, and less disruptive to existing organizational structures.  Their results are also easier to quantify, making them more attractive from a traditional return-on-investment (ROI) perspective.

Strategic capabilities, by contrast, address core decision-making and long-term performance.  These applications are designed to improve how the organization competes, allocates capital, selects markets, or manages portfolios.  Examples include market selection models, asset allocation systems, portfolio optimization tools, and predictive analytics for development feasibility.  Strategic capabilities often span multiple functions/departments and influence decisions at a firm-wide level.  They are usually more complex, require more resources, and often involve rethinking existing workflows. Unlike operational tools, strategic AI initiatives aim to create competitive advantage, not simply efficiency.

Key Differences Between Operational and Strategic Capabilities

Operational and strategic capabilities differ along several dimensions, including implementation complexity, cost, risk, and impact.  Operational initiatives are generally faster to implement and less disruptive.  They often build on existing workflows, automating tasks that employees already perform manually. Because of their incremental nature, they can be deployed with minimal disruption to current processes and low leadership involvement.  These characteristics make operational projects appealing as early AI initiatives.

Strategic capabilities are inherently more disruptive.  They often require new data infrastructure, new skill sets, and significant changes to decision-making processes.  Leadership involvement becomes critical because strategic systems influence how capital is allocated and how performance is evaluated. These initiatives are also riskier, as they frequently attempt to accomplish tasks that have not previously been implemented within the organization.  As a result, they require greater investment, more risk tolerance, and longer time horizons.

Another important difference lies in measurability.  Operational capabilities typically produce outcomes such as reduced labor time, faster reporting cycles, or fewer errors.  These are all things that are relatively easy to track and do a “before and after” comparison.  Strategic capabilities, however, generate value through better decisions, which are fundamentally harder to quantify.  Developing a system that produces more accurate forecasts for growth rates in different markets could lead to changes in how the firm allocates capital, and thus the returns the firm generates.  But how much more yield?  And over what time horizon?  More cash flow or growth in value?  These are harder things to measure.  The impact may appear indirectly through improved portfolio performance, better risk management, or superior market positioning.  This makes strategic investments more difficult to justify using traditional ROI frameworks (I discuss the problem strategic applications pose for traditional ROI analysis on Friday).

The Capability Spectrum

Although operational and strategic capabilities are conceptually distinct, they exist along a continuum rather than as discrete categories.  Some applications fall clearly into one group, while others fall in the middle or on both sides depending on the focus of the AI initiative.

Let’s look at acquisitions analysis for a multifamily investment firm as an example.  Depending on what task you choose, the outcome could be operational or it could be strategic.  Operational tasks might be automating the extraction of data from sources that are now extracted manually by analysts and then automatically populated in an underwriting template.  That’s a relatively easy thing to measure.  How much time/cost does it take to move that data now, how much will it cost to automate that process, and how much time/cost will it take to perform that same task after the automation is put in place? 

However, the same acquisitions analysis function could be a strategic use case if it involves improving the depth of analysis and insight executives have at their disposal.  Machine learning and deep learning applications can provide significantly more insight into potential growth rates of different markets, economic or demographic trends within markets and submarkets that make one property or property type more attractive than another over a 5-year hold period, or segmenting of markets by near-term risk and uncertainty.  This kind of insight allows executives to make decisions about where to focus acquisitions and increasing the confidence in capital allocations.

Ideally you would do both.  Automate as much as possible and analyze as much as possible.  You would just need to be clear about the process, the resources, and the overall metrics for success.

Similarly, development feasibility tools may begin as operational tools but transition into strategic decision-support systems as they incorporate predictive modeling and portfolio optimization.  And there are thousands more examples that illustrate the same tradeoff. 

So just saying that you want to improve acquisitions analysis is extremely ambiguous.  What part do you want to improve?  What does improving each part involve (simple automation or the development of machine learning infrastructure and algorithms)?  What resources will be needed for each (skillsets, time, costs, etc.)?  What is the desired outcome (efficiency or competitive advantage)?  How much risk are we willing to tolerate? 

The answers to these questions will be vastly different depending on what you want to accomplish.  Using the same evaluation and decision-making framework for both categories doesn’t make sense. 

This kind of ambiguity in proposed projects makes evaluation challenging.  Misclassifying initiatives can lead to unrealistic expectations.  For example, incremental automation may be labeled as “transformation,” leading leadership to expect competitive advantages that the system cannot realistically deliver.  Just automating data extraction isn’t going to give you better decision-making capabilities.  This mismatch often results in disappointment and contributes to “proptech fatigue,” where firms become skeptical of AI initiatives after early projects fail to meet inflated expectations.

Strategic Caveats

Strategic capabilities are admittedly more difficult and come with several important caveats.  I think this is one of the main reasons we haven’t seen more strategic technology applications in real estate. First, they are inherently abstract. Unlike operational tools, which should produce efficiency gains that are easy to measure, strategic systems influence decision-making indirectly.  This makes them harder to evaluate and more difficult to communicate within the organization.  Leadership must be willing to accept some uncertainty and longer time horizons.  This often starts with leadership getting educated on technology and AI, in particular.  We find that most executives in real estate have limited exposure to technology development concepts, significantly limiting their ability to provide guidance and make decisions for technical initiatives.

Second, strategic initiatives require creativity.  Organizations must identify opportunities that do not yet exist and envision new ways of operating.  This requires both domain expertise and technical understanding.  Without this combination, firms may struggle to define meaningful strategic objectives.  This is also the place where the, “that’s not how it’s done” roadblock emerges.  This can be extremely frustrating for teams when the entire purpose of developing strategic capabilities is to do things differently (hopefully better) than they’re currently done using new tools and new approaches. 

Side note:  New capabilities come from new tools.  As mentioned, machine learning and deep learning models provide much deeper analysis of markets.  These new tools allow real estate to ask questions it simply didn’t have the tools to get answers to before.  An analogy I like to use is this:  If you’re a carpenter and you only have one tool at your disposal, you’re limited in what you can build.  But if someone brings you an entire workshop of new tools, you’ll be able to build more and more complex things.  The “that’s not how it’s done” stems from not understanding these new tools and lead to limitations in firms that are much more restrictive than they should be.

Third, strategic initiatives require significant leadership buy-in.  Because they are disruptive and resource-intensive, they cannot be implemented solely within isolated teams. Leadership must support changes to workflows, performance metrics, and organizational structure. Without this support, strategic projects often stall or fail to achieve meaningful adoption.

Another important caveat is that improving existing processes may not always be the optimal strategy.  In some cases, AI enables entirely new approaches that replace traditional workflows.

For example, instead of attempting to make the market analysis process more efficient with better automation for the analysts, machine learning applications could replace a large part of that process altogether. 

Organizations must remain open to rethinking how they operate rather than simply optimizing current methods.  This can be particularly challenging for leadership, as decisions about cutting-edge technology often fall to individuals with limited technical backgrounds (again emphasizing the importance of decision-makers improving their AI knowledge).

Operational Capabilities as a Prerequisite for Strategy

Although strategic capabilities create the greatest potential for long-term value (in my opinion), they often depend on operational foundations.  Machine learning systems require structured data, and structured data typically results from operational digitization and automation.  If an organization’s data is fragmented across spreadsheets, documents, and manual workflows, strategic analytics become difficult or impossible.

Operational efficiency initiatives frequently involve digitizing internal processes and standardizing data.  This digitization creates a pipeline for collecting and organizing information.  Once data is available in a consistent digital format, it can be used for advanced analysis.  In this sense, operational capabilities serve as a prerequisite for strategic capabilities and have an additional added value.

However, organizations must avoid treating digitization as the final objective. Many firms digitize data without considering how it will be used for analysis. As a result, data may be stored in formats that are unsuitable for machine learning. Digitization should therefore be viewed as a means to enable strategic analysis, not as an end goal.

Conclusion

Operational and strategic AI capabilities serve different but complementary roles within real estate organizations.  Operational tools improve efficiency and create the data foundation required for more advanced systems.  Strategic capabilities leverage that data to improve decision-making and generate competitive advantage.  While operational initiatives are easier to implement and measure, strategic initiatives offer significantly greater long-term value.

Organizations must therefore balance both approaches.  Overemphasis on operational automation may lead to incremental improvements without meaningful differentiation.  Conversely, pursuing strategic initiatives without operational foundations may result in failure due to insufficient data and infrastructure.  A deliberate progression from operational digitization to strategic decision support offers the most effective path for realizing the full potential of artificial intelligence in real estate.

Getting clear about the outcomes, resources, and direction the company wants to go is the first, and I think most important, step in the evaluation process for potential AI applications.  I’ll talk more about how much this distinction matters when we get to ROI on Friday and bring together the 1) operational vs. strategic, 2) value, technical, and execution feasibility, 3) sources, and 4) ROI evaluation at the end of this series.

Tomorrow I’ll explain how we think about AI feasibility, including the value a project will bring (should we do it), whether it is technically feasible (can it be done), and if it is technically feasible, does your firm have the resources and ability to tackle it (can WE do it).