In many ways, site selection has become a finely-tuned process. Data on key variables -- labor market trends, real estate, transportation and energy costs, and business climate rankings -- are carefully considered. Site selection teams assess operational risks including geopolitical conditions, logistics dependability and construction delays. Economic incentive negotiations elicit millions of dollars from government entities.
The best site selection processes go even further. Often, project teams look at site characteristics individually and overlook how possible sites fit with other locations in the company’s portfolio and with supplier and customer locations. Whether looking for a new site or an expansion, whether new space is required for manufacturing, distribution, customer contact centers, back office activity or other office and R&D space, site selection requires a portfolio perspective.
It’s as if there’s a bike race in which each racer is on his or her own. Bike racers know that they are more efficient working as part of a team. Similarly the typical corporation has a team of properties. Finding the best location means understanding how a new team member fits with the rest of the team.
Outsourcing for more capacity
Site selection from a portfolio perspective begins by considering whether new capacity is really needed. Each new location or expansion is an investment that’s difficult to undo -- valuable capital resources are locked up for years. Many companies don’t consider what will happen when the economy slows, as is almost guaranteed.
One alternative to adding capacity is outsourcing or a strategic partnership with third party providers. Typically, outsourcing is considered on an all or nothing basis, with the recommendation that companies outsource non-core activities to achieve lower costs, better business processes and improve focus on core activities. Outsourcing activities on a partial basis is another alternative. Outsourcing can improve flexibility and enable companies to respond more quickly to changing business conditions. Outsourcing contracts require shorter term commitments than owning or leasing a property. Third party providers are responsible for adding or reducing staffing, which can be shifted among different clients.
Consider Walmart -- much of its success is based on its incredibly efficient distribution system, which means distribution is one of Walmart’s core competencies. In line with standard theory, Walmart owns and operates 97 distribution centers in the US. But they also have 18 distribution centers operated by third-party operators which provide flexibility.
Excess capacity in other parts of the organization
Many companies operate with business unit silos, unaware that other business units have excess space and labor resources that would be suitable for a growing business unit’s expansion. Rather than paying for new space, why not use space that the company is already paying for?
In the past, GE’s business units operated their real estate activities independently. One business unit would be looking for new space at the same time that another business unit had excess space in the same area. But no one knew where this excess space and resources were. To identify opportunities, GE implemented a major corporation-wide real estate database. The initiative was so important that GE’s CEO Jeffrey Immelt cited “fewer rooftops” as one of GE’s cost reduction strategies in his 2006 Harvard Business Review interview.
Senior executive support and encouragement, such as Immelt’s in this case, is key to the success of these initiatives. Business managers often believe that they have their own special real estate requirements. In some cases this is true, and alternative location may reduce productivity more than costs. But often, it’s a case of corporate politics and senior executive direction is needed.
Another reason GE can consider collocating businesses is its long-standing focus on “bottom-line discipline.” This real estate strategy was part of Immelt’s charge to reduce general and administrative costs as a percent of sales from 11% to 8%. Even as leaders in their industries with outstanding profit margins, GE business leaders understand that they always need to be looking for ways to reduce costs.
Collocation not only reduces current costs, it will also reduce future costs. In the traditional silo approach, each business unit takes on enough space to accommodate its high growth target. But looking across the portfolio, it’s unlikely that all business units will grow at a high rate. With business units located in one facility or in a cluster of properties, growing business units can absorb space as others need less space. Capacity can be based on the average rather than high growth forecasts. As Figure 1 shows, clustering properties so that space could be shared reduced real estate costs by 8% for one client.
At Intel, the corporate real estate and site development department avoids business unit silos by convening cross-functional teams for each property type, including manufacturing, research and development, and back office, notes Jack Moretti, Intel’s enterprise strategic planning manager. These teams, which include the real estate department’s customer relationship managers and representatives from each relevant business unit, meet on a regular basis to discuss new and changing business requirements and how these needs can be coordinated.
Portfolio considerations when adding capacity
When additional space is required, a portfolio perspective to site selection means companies should evaluate how new locations or expansions fit within the rest of the portfolio. For manufacturing and distribution locations, how would each of the possible locations complements existing locations in their access to customers and suppliers. A manufacturing location might have lower labor, energy and real estate costs but the local markets may already be served by another location or transportation costs may be higher due to a greater distance to suppliers and customers.
For customer contact centers and back office operations, the analysis should consider how the site will improve diversification across weather patterns, political risks, changing demographics and time zones. Are all of the locations in the same tornado alley? Are service agents distributed across the country in proportion to when customers are likely to call?
At Intel, another portfolio consideration is how a location contributes to Intel’s “strategic presence.” Each site provides valuable marketing exposure and community goodwill notes Sharon Barshaf, Intel’s site selection program manager. Intel enhances its strategic presence by locating sites throughout its customer base.
Site selection activities can be prompted be situations other than the need to increase capacity. Portfolio realignment can be beneficial after mergers and acquisitions, in portfolios that resulted from individual, uncoordinated decisions, and in response to a changing customer base or changing input costs and availability. For example, recent increases in energy costs are prompting many companies to rethink manufacturing and distribution locations to site locations closer to customers and raw materials.
Unilever’s Home and Personal Care division realized they could reduce costs and improve delivery times following its acquisitions of Lever Brothers, Chesebrough-Pond''s and Helene Curtis. After these acquisitions, Unilever HPC had fifteen distribution centers with overlapping service areas. Furthermore, major customers such as Walmart and Kmart were demanding lower costs and more efficient delivery schedules. Their new distribution network could include either expansion of existing distribution centers, completely new sites, or both.
Intuition or systematic analysis?
With demanding and cost-conscious customers, Unilever understood that they needed to make sure that they were selecting the most cost-effective set of locations but they knew a comprehensive portfolio analysis would be complex and time-consuming. Like a game of pickup sticks, removing or adding one location would affect distribution activities through out the portfolio. There were so many different alternative combinations, they could easily overlook the most cost-effective solution.
To address this complexity, Unilever and its strategic partner Prologis used advanced analytical techniques supported by specialized software, in this case INSIGHT’s SAILS supply chain and logistics optimization software. In this approach, analysts provide the traditional site selection costs for existing locations and for any possible new locations. Then the software’s mathematical algorithms search through all of the different combinations of locations to find the one that meets the business needs for capacity, delivery times and other factors at the lowest cost.
With this approach, the software can search through the alternatives in minutes. In contrast, without these systems, the analysis can take days, weeks, even months and there is no assurance that analysts aren’t overlooking a superior solution. This approach is easier, faster, and can find millions of dollars in savings. It’s the difference between having a good solution and hitting the bulls eye.
After considering solutions with different opportunities for cost savings and improved delivery schedules, the project team selected a solution with five mega distribution centers, three were expansions of existing centers near Atlanta, Dallas and Carlisle Pennsylvania and two were selected from possible new locations, one near Saint Louis, Missouri and the other near Los Angeles. The new portfolio provided $20 million in annual savings and improved delivery schedules for major retailers.
Portfolio optimization for other activities
While these portfolio optimization techniques are most commonly used for manufacturing, warehouse and distribution analysis, they can also be applied in other areas. At Critical Core, we have extended these techniques to evaluate customer contacts centers, back office activities and regional portfolios of office and R&D locations. These techniques can also be applied to areas that don’t typically lead to major site selection searches such as locating retail branches, service centers, and schools.
In our work, we have found that the benefits go beyond cost savings and increased efficiencies to better community goodwill. One of our clients, a large financial services company, recognized that it could save millions by consolidating their back office activities located in over twenty cities. They found several solutions using the traditional jigsaw puzzle approach, in which planners pick possible locations using intuition and then calculating labor and real estate costs. Then they asked us to evaluate the portfolio. One of the other solutions we found improved cost savings “slightly,” another $20 million over the analysis period. But more importantly for community goodwill, these savings could be achieved with one-third fewer layoffs because our solution found more ways to retrain current workers for new activities.
Combining a portfolio perspective with long-term planning
Considering future changes, such as changing labor markets, demographics and business conditions, also requires a portfolio evaluation. In one case, the real estate department of a large financial services firm reviewed its portfolio and found one division had all of its call center operations in high-wage locations. The real estate department recommended diversifying some operations to a lower wage location and collocating them with call center activities for other divisions. The business leaders rejected this proposal. Its current competitive advantage provided additional revenues to cover higher labor costs, but what would happen as competition eroded the competitive advantage and drove them to look for a lower cost location. From a forward looking perspective, we emphasized that by starting to transition jobs to a lower cost location through natural turnover, they could avoid laying off hundreds of employees and preserve community goodwill.
Although many of the Fortune 500 use advanced analytical techniques occasionally to optimize their manufacturing and distribution networks, probably less than 25% use them consistently or extensively estimates INSIGHT’s Tony Mirra. Adoption for customer contact centers, back office activities, and regional planning is considerably less as is adoption in smaller companies. With proven results, portfolio optimization experts agree that many more organizations could find savings and improve efficiency by using these techniques. Often, business managers are unaware that these techniques exist or may be put off by technical jargon -- the manufacturing and distribution analysis techniques are often referred to as “strategic supply chain network optimization.”
As well, business managers would often rather live with a suboptimal solution pieced together in a spreadsheet than live with a solution they don’t understand. To many, these techniques seem like black boxes - data go in and results come out. But if we avoided black boxes, we would miss the benefits of computers , new cars with sophisticated safety systems, and Google. The “garbage in, garbage out” phenomena is another concern. Yet the data used in these systems is the same data that is used in most traditional analyses. With these techniques, the analysis is more systematic, comprehensive and analysts can complete sensitivity tests and test alternative forecasts.
In our increasingly cost-conscious and competitive business environment, companies no longer have the luxury of relying on individual site selection evaluations. They can find better solutions by looking at alternatives from a portfolio perspective, considering how possible new locations fit within their existing portfolio and the locations of their suppliers and customers. And on an on-going basis, they should reassess their portfolios to see if changing business conditions provide opportunities for even more savings and efficiencies. Finally, companies don’t need to be overwhelmed by the complexity of these decisions. Proven portfolio optimization techniques exist to make these analyses easier and faster, with even better solutions.