Corporate Real Estate alignment: A preference-based design and decision approach
One of the long-standing issues in the field of corporate real estate management is the alignment of an organization’s real estate to its corporate strategy. In the last thirty years, fourteen Corporate Real Estate (CRE) alignment models have been made. In some of these CRE alignment models it is indicated that they strive for maximum or optimum added value. Even though extensive research into these existing CRE alignment models has provided us with valuable insights into the steps, components, relationships and variables that are needed in the alignment process, these models still fall short in two ways. Most models pay little to no attention to
1 The design of new CRE portfolios;
2 The selection of a new CRE portfolio that adds most value to the organization.
How a CRE manager is able to design and select an optimum alternative in an operational way remains a black box in many alignment models.
In CRE alignment models, the authors generally use either the stakeholder or the shareholder approach. Both approaches received criticism in the past. Kaplan and Norton (2006) state that the shareholder approach with purely financial measures of performance are not sufficient to yield effective management decisions. Jensen (2010) criticizes the stakeholder approach and states that managers in an organization need to define what is better and what is worse which forms the basis of making decisions. In his view, putting them in opposite positions is not correct because both are of a different nature. In fact, Jensen (2010, p. 33) states “ ... whether firms should maximize value or not, we must separate two distinct issues;
1 Should the firm [organization] have a single-valued objective?;
2 And, if so, should that objective be value maximization or something else ...?"
I agree with Jensen’s view that a single-valued objective function is needed, but argue that in our CREM domain a financial measure is not fully suitable. A financial measure is not suitable, because values (also referred to as qualities) of buildings fall in two general categories.
These categories are often interrelated and overlap in practice as explained by Volker (2010, p. 17), the categories are:
—— “technical, physical, hard, functional, objective or tangible qualities;
—— perceptual, soft, subjective, judgmental or intangible values.”
These intangibles are vital to CRE management but often suppressed. Real estate decision making therefore needs to be able to include all of these values in order to be purposeful. If they are treated separately, the restriction is that one effect can be more difficult to monetize than the other effect, as shown by Mouter (2012) and if multiple measures are used as in the stakeholder approach ”if you take one set of quantifiable impacts and one set of non-quantifiable impacts in an appraisal, one set will dominate” (Mishan, in Mouter, 2012, p. 10).
Research aim: The aim of this research is to enhance CRE alignment by improving CRE decision making in such a way that corporate real estate managers are able to determine the added value of a particular corporate real estate strategy quickly and iteratively design many alternative real estate portfolios.
Conclusions about developing the Preference-based Accommodation Strategy design and decision approach
This research successfully developed, tested and evaluated a new design and decision approach in corporate real estate alignment that makes it possible to design alternative CRE portfolios and then to select the portfolio that adds most value to the organization. The originality of this research to (1) define value as technically equivalent to preference and (2) use a design and decision approach for the alignment problem. This new approach is called the Preference-based Accommodation Strategy design and decision approach (PAS). PAS was developed and tested in accordance with the five stages of an operations research project. PAS is constructed upon fifteen basic concepts and definitions from management science, decision theory and design methodology.
Preference Measurement and Preference-Based Design are the most important basic concepts. By using the overall preference (value) score as overall performance measure, based on a single-valued objective function, CRE managers are able to select a new CRE portfolio that adds the most value to the organization. Following Barzilai (2010), all tangible and intangible values are categorized either as physical or nonphysical properties of an object. To enable the application of mathematical operations to these non-physical properties, such as preference, Barzilai (2010) developed a theory of (preference) measurement as well as a practical evaluation methodology Preference Function Modeling for constructing proper preference scales. To enable the design of alternatives the Preference-based Design method (Binnekamp, 2011) is used as particular technique in the domain of design and decision systems. By adjusting this method it can be used on portfolio level.
PAS is structured around three decision making rationalities (Kickert, in De Leeuw, 2002). The three components are; the steps (procedural rationality), the stakeholders & activities (structural rationality) and the mathematical model (substantive rationality) as shown in Figure S.1. The substantive rationality enables the decision maker to choose an alternative based on the bounded rationality perspective. The procedural rationality enables the decision maker to take into account the time perspective when selecting an alternative and the structural rationality enables that more than one decision maker is involved. By using all concepts past experience has benefited the development of PAS. For PAS to be operational all components are connected coherently.
The coherence between the components is shown in a flowchart in Figure S.2. In the steps, decision makers define decision variables representing accommodation aspects that make the accommocation stratgy tangible and iteratively test and adjust these variables by designing new alternative real estate portfolios. The alternative design that adds most value to the organization, i.e. has the highest overall preference score, is the portfolio that optimally aligns real estate to corporate strategy. The activities that the participants perform are a series of interviews and workshops, while the system engineer builds the accompanying mathematical models. The approach overcomes the problems inherent to the current models and uses explicit scales for measuring preference, i.e. value, defined by stakeholders themselves.
Conclusions about testing PAS
PAS is tested successfully in three pilot studies. All pilot studies show that the stakeholders were able to perform all the steps and activities, including the steps to determine preference curves (step 2) and the design alternatives themselves (step 5). The stakeholders were able to design an alternative CRE portfolio with a higher overall preference than in the current situation Table S.1. An added value of 54, 17 and 5 (out of a 100) was achieved either by the stakeholders (in step 5a) or the optimization tool (in step 5b). In the last step, all stakeholders accepted that alternative as the final outcome. Next to that, there is an indication, based on the third pilot study, that the use of the preference curves in PAS improved the representation of the stakeholders preferences than in their current scorecard system.
In the first and third pilot, alternative CRE portfolios have been generated with an optimization tool (step 5b). Due to the nature of third pilot the brute force approach was used successfully in generating a global optimum (see Table S.1). In the first pilot, the algorithm (step 5b) was not able to generate a local optimum because a subset of the alternatives was infeasible. The feasible set of alternatives could not be characterized mathematically and was not available to the algorithm. The brute force approach is preferable to the search algorithm as it finds a global optimum instead of a local optimum but has as disadvantage that it often cannot be used when a pilot is too complex. In PAS, stakeholders design alternatives (step 5a), and use the PFM algorithm to rate them as has been done for the first two pilots.
Conclusions about evaluating PAS; iteration is the key
In all three pilots the stakeholders as well as the observers evaluated PAS very positively. According to the stakeholders, determining preferences and refining and adjusting them in collective workshops is the attractive part of PAS. The participants indicated that, whilst the method of determining preferences is easy, accurately determining which preference is related to a certain decision variable value is not.
Assigning preference scores to decision variable values can be arbitrary at first. By repeating the cycle of determining preferences and making designs a number of times, the stakeholders see the effect of the decisions made in the design, and how their preferences affect those decisions. In all pilot studies the decision makers used the opportunity to either add or remove decision variables and change curves, weights or constraints. The use of such a learning process in the context of work practice and problem solving is described by Schãn (1987) as reflection in action.
Conclusions about reflecting upon PAS
PAS as design and decision approach can be used as add-on to existing CRE alignment management models. However, using PAS as add-on in these models creates methodical difficulties. The structure of these models is often not congruent with the PAS structure. To avoid these difficulties, PAS is also described both from a systems’ management perspective (De Leeuw, 2002).
The three pilot studies showed that PAS can be applied in different organizations, and for different types of problems with a different level of complexity. In comparison, the first two pilots were more complex because more stakeholders were involved and more interventions were possible. Applying this approach to multiple context-dependent cases has yielded more valuable results than just applying it to one case. Based on the results of this study, it is justified that PAS can be used for a wide range of real estate portfolio types.