Warehouse and Distribution Center Site Selection
A network model says roughly where a node belongs. Site selection decides which building, and it turns on inputs of unequal quality. Labor and power data come from government sources.
Labor now leads the decision. A freight-optimal site in a market that cannot staff it is not optimal, and the labor data is independent and metropolitan-level.
Power can be the critical path. For automated facilities and fleet charging, interconnection timelines can exceed the entire project schedule and should be checked before a shortlist is made.
The shell forecloses the automation. Clear height, column spacing, floor flatness, dock ratios, and trailer parking determine what can be installed later, and they cannot be changed afterward.
The market data has no independent publisher. Triangulate across sources, disclose the interest, and anchor the decision on the labor and power data instead.
Incentives carry clawbacks. The headline value matters less than the conditions attached and what happens if headcount targets are missed.
Market overview
The short answer
Site selection is the step after network design. The model identifies roughly where a facility should sit to serve demand at acceptable cost; site selection chooses the actual market, the actual building or parcel, and whether to lease or build. Three things dominate the decision in current conditions. Labor availability frequently outranks transportation cost, because a site that cannot be staffed cannot operate whatever its freight math says. Power availability has become a genuine constraint for automated facilities and fleet charging, with interconnection timelines measured in years in some regions. And the market data everyone quotes, meaning vacancy, absorption, and asking rent, has no independent publisher: it comes from commercial brokerages and a data vendor, all of which earn revenue from the transactions and subscriptions concerned.
KEY FACTS
Verified August 2026. Each statement below is complete on its own and cites its source in section 08.
How do we get from the network model to a building?
The sequence has four stages and they narrow quickly. The network model produces a geography: a region or a set of candidate metropolitan areas that serve the demand profile at acceptable cost and service. SCR covers that modeling in its network design guide and this page begins where it ends. The second stage is market selection, which tests those candidate areas against the constraints the model does not carry: whether the labor exists, whether power is available on a workable timeline, whether suitable buildings exist at all, and what the regulatory and community environment is.
Figure 1. The funnel, with each decision input colored by the provenance of its underlying data. Labor and power rest on government and national laboratory sources. Vacancy, rent, and absorption rest on parties that earn fees from the transactions they report, which is the asymmetry this page exists to make visible.
The third stage is building selection within a chosen market, where the specification questions in section 05 dominate, and the fourth is the transaction: lease, build to suit, or acquire. The lease or build decision turns on speed, capital availability, how specific the requirement is, and how long the operation expects to occupy the site. A build to suit produces exactly the required building on a longer timeline; an existing building is faster and imposes its previous occupant's assumptions on the operation.
One sequencing error is common and expensive: shortlisting buildings before testing labor and power. Both are market-level constraints rather than building-level ones, and discovering after a shortlist that the metropolitan area cannot supply staff, or that the utility cannot supply capacity within the project window, means restarting rather than substituting. Test the market constraints first and the buildings second.
How do we analyze a labor shed?
A labor shed is the geographic area from which a facility can realistically draw workers, defined by commuting patterns rather than by administrative boundaries. Analyzing one means asking how many people of the required skill profile live within an acceptable commute, how many are already employed doing similar work nearby, what they are paid, and how contested that pool is by other employers.
Unusually for this decision, the data is independent and good. The Quarterly Census of Employment and Wages covers over ninety-five percent of United States jobs and reports employment and wages by industry at county and metropolitan level, including warehousing and storage, which allows a direct measure of how large the existing warehouse employment base is in a candidate market and how fast it has grown. The Occupational Employment and Wage Statistics program adds occupational detail and metropolitan wage estimates for the specific roles a facility will hire. Job openings and turnover data provides regional context on how tight the market is.
Two analytical cautions matter. A large existing warehouse employment base is ambiguous evidence: it proves the labor exists and also indicates the pool is already contested, and a market with several large operators competing for the same profile can be harder to staff than a smaller one with none. And wages in the published statistics are a floor rather than a forecast, since an entrant adding significant demand to a tight market will pay above the current median, which is a cost the model should carry rather than a surprise the operation absorbs.
Turnover deserves explicit attention because it is the cost that persists. A site that can be staffed at opening and cannot retain staff carries recurring recruitment and training cost and a permanent productivity penalty from an inexperienced workforce. National turnover statistics from staffing firms and industry associations circulate and are interested-party material; the government job openings and turnover series provides regional context, and a candid conversation with operators already in the market provides the rest.
Is power the new critical path?
For a conventional facility, power is a utility arrangement handled in the ordinary course. For an automated facility, and for any operation planning to charge an electric fleet, the required capacity can be large enough that it becomes a constraint on the project schedule rather than a line item within it. The question is not whether power is available in the abstract but whether the utility can deliver the required capacity to that parcel within the project window.
The independent evidence on how slow this can be comes from the national laboratory that tracks interconnection queues. Its analysis reported that projects reaching commercial operation in 2024 had spent an average of about fifty-five months in the queue, with roughly 2,290 gigawatts of capacity active in queues at the end of that year across some ten thousand projects. In one large regional market, the total wait from application to commercial operation rose from under two years in 2008 to over eight years by 2025. Federal regulators have restructured the process, replacing sequential studies with cluster studies, and the backlog remains substantial.
Two qualifications keep this accurate. Those figures describe generator interconnection, meaning parties connecting generation or storage to the grid, which is a related but distinct process from a large new load connecting to a distribution or transmission system. Large load interconnection is itself under active regulatory attention as data centers and electrified fleets change demand patterns. The relevant point for a site decision is directional rather than numeric: in constrained regions, capacity delivery timelines can exceed a facility development schedule, and the only way to know is to ask the specific utility about the specific parcel early.
The practical instruction follows. Put a capacity question to the utility during market selection rather than during building selection, state the expected load including any charging, and get the answer about that location in writing. A site that cannot be powered on the required timeline should leave the shortlist before anyone tours it.
Which building specifications constrain automation?
Five specifications determine what an operation can do inside a building, and none can be changed materially after the fact. Clear height is the usable vertical dimension beneath the lowest obstruction, and it sets the ceiling on storage density and on whether tall automated systems fit at all. Column spacing determines the usable floor grid: closely spaced columns obstruct racking layouts and constrain the paths mobile robots can take. Floor flatness, specified through defined tolerance measures, matters more than it sounds because tall narrow-aisle equipment and some automated systems have flatness requirements that an ordinary slab does not meet.
Dock door ratio, meaning doors relative to floor area, determines throughput capacity at the interface with transport. A building with too few doors for the volume becomes constrained at the dock regardless of what happens inside it, and adding doors to an existing structure is possible and expensive. Trailer parking is the specification most often underestimated: an operation running drop trailers needs parking positions, and a site without them will either turn away trailers or park them on public roads, which creates a community relations problem alongside an operational one.
Table 1. The five specifications. The third column is the one to weigh during selection: two of these cannot be remedied at any price, which means the building chosen today sets the automation ceiling for the whole occupancy.
The strategic point is that these decisions are made once. An operation that intends to automate within the occupancy should specify for the automation it might install rather than for the operation it runs on day one, because the marginal cost of a taller, better-configured building at selection is far below the cost of discovering three years later that the intended system does not fit.
How do we read market data, incentives, and zoning?
Start with provenance, because this is where SCR's advice differs most from the standard treatment. There is no federal or independent publisher of industrial vacancy, net absorption, or asking rent. The only federal vacancy product covers residential housing. Every industrial market figure in circulation originates either with a commercial brokerage, which earns fees on the transactions it reports, or with a for-profit data vendor selling subscriptions, whose own regulatory filings confirm that clients use its product to calculate exactly these metrics. Landlords and industrial property owners publish their own figures and have a direct interest in a rising rent narrative.
This is not an argument for ignoring that data, because no alternative exists and a real site search cannot wait for one. It is an argument for handling it differently: triangulate across several brokerages rather than relying on one, present ranges rather than point figures, disclose the commercial interest in any internal paper that uses it, and anchor the decision on the labor and power data where independent sources do exist. A recommendation that rests on a single brokerage's vacancy figure rests on a number with a conflict attached.
Economic development incentives are the second area requiring care. The headline value is the least important term. What matters is the conditions: what headcount, wage level, and capital investment are required, over what period, verified how, and what happens if the operation falls short. Clawback provisions can require repayment of benefits already received, and an operation that automates more than it forecast, and therefore employs fewer people, can trigger one by succeeding operationally. Model the incentive against the operating plan, including the automation roadmap, rather than against the headcount in the application.
Zoning and community opposition has become a real constraint rather than a formality in several United States regions. Peer-reviewed planning research examining recent zoning actions on warehousing found discretionary permitting responses, including denials, moratoriums, and conditional use requirements, to be the most common local response, and multiple jurisdictions have adopted moratoriums on new warehouse development. Two practical consequences follow: a parcel zoned appropriately today may still face a discretionary approval that can be refused, and the traffic and trailer parking questions in section 05 are frequently the specific grounds on which opposition succeeds.
The fair case against this page's skepticism about market data deserves stating. Brokerage and vendor data, whatever its conflicts, is the only source granular and timely enough to run an actual site search: no government series publishes submarket vacancy or asking rent, and refusing to use it means having no market view at all. That is correct, and the position here is not a boycott. It is that the data should be triangulated, its interest disclosed, and the weight of the decision placed on the inputs where independent evidence exists.
Frequently asked questions
Does labor or transportation cost matter more now?
Labor frequently outranks transportation cost, because a site that cannot be staffed cannot operate whatever its freight math shows. Transportation cost remains a real optimization, and it is the network model's question. Site selection is largely a labor and power question.
What government data exists on warehouse wages and employment?
The Quarterly Census of Employment and Wages reports employment and wages by industry at county and metropolitan level, covering over ninety-five percent of United States jobs, including warehousing and storage. The Occupational Employment and Wage Statistics program adds occupational detail with metropolitan estimates.
What is a labor shed?
The geographic area from which a facility can realistically draw workers, defined by commuting behavior rather than administrative boundaries. Analyzing one means measuring the available pool of the required profile, what it is paid, and how contested it already is by other employers.
How do I estimate turnover in a market?
Use the government job openings and turnover series for regional context and speak to operators already in the market. National turnover figures published by staffing firms and industry associations are interested-party material and should not be treated as benchmarks.
How long does electrical interconnection take?
It varies enormously by region. Independent national laboratory analysis reported an average of roughly fifty-five months in the queue for projects reaching commercial operation in 2024, and over eight years from application to operation in one large regional market. Ask the specific utility about the specific parcel.
What clear height should a modern facility have?
It depends on the intended storage system, which is the point: specify for the automation you might install rather than for the operation you run at opening. Clear height and column spacing cannot be changed later, so they set the ceiling for the whole occupancy.
Is there an independent source for industrial vacancy and rent?
No. The only federal vacancy product covers residential housing. Industrial figures come from commercial brokerages that earn fees on the transactions they report, or from a for-profit data vendor selling subscriptions. Triangulate across sources and disclose the interest.
What is a clawback provision?
A term in an economic development incentive requiring repayment of benefits if the recipient fails to meet its commitments on headcount, wages, investment, or duration. An operation that automates more than forecast can trigger one, so model the incentive against the automation roadmap.
How real is community opposition to new warehouses?
Real enough to be documented in peer-reviewed planning research, which found discretionary permitting responses, including denials and moratoriums, to be the most common local reaction to warehouse development. Several jurisdictions have adopted moratoriums outright.
Should we lease or build?
Build to suit produces exactly the required specification on a longer timeline and with more capital commitment. Leasing an existing building is faster and inherits the previous occupant's assumptions. The more specific the operational requirement, and the longer the intended occupancy, the stronger the case for building.
Method, sources, and where to go deeper
Method
Labor evidence follows the United States Bureau of Labor Statistics directly, including the employment and wage census and the occupational wage program, both of which publish at metropolitan level.
Interconnection evidence follows a national laboratory's published queue analysis and the federal energy regulator, with the distinction between generator and large load interconnection stated explicitly.
Zoning evidence follows peer-reviewed planning research rather than trade commentary.
Market data provenance was checked against the data vendor's own regulatory filings and against the federal statistical agencies, to establish what independent sources exist rather than assuming.
Supply Chain Research is independent and vendor-neutral. We accept no payment from the vendors or categories covered, and this page names no products.
Caveats
SCR publishes no benchmark for industrial vacancy, absorption, or asking rent, because no independent source produces one. Figures in circulation come from commercial brokerages and a for-profit data vendor, all of which have a commercial interest in the market they report on.
The interconnection figures cited describe generator interconnection queues, which are related to but distinct from large load interconnection. They establish that timelines can be long in constrained regions and should not be read as a forecast for a specific site.
The wage figure quoted is a national median for a broad occupational category at a stated date. Metropolitan and occupation-specific figures should be taken from the wage statistics program directly.
Counts of zoning actions circulating in secondary coverage of the planning research should be verified against the article itself before being repeated, as they are easily conflated with other figures.
Figure 1, Table 1, and Table 2 are structural summaries rather than measured research findings. Nothing on this page is real estate, tax, or legal advice.
Where to go deeper
Readers whose question is where facilities should sit rather than which building to take should read the SCR guide to supply chain network design, which owns the modeling this page begins after. The warehouse automation and robotics guide covers the systems whose requirements drive the specifications in section 05, and the WMS, WES, and WCS guide covers the control software. The yard management guide covers trailer parking and gate operations, which recur in both the operational and the community objection. Readers scoping across categories should start with the SCR supply chain software category map.
Sources
Sources
- US Bureau of Labor Statistics. Quarterly Census of Employment and Wages. Primary government source. County and metropolitan employment and wages by industry.
- US Bureau of Labor Statistics. Warehousing and storage industry statistics. Primary. Industry-level employment and wage data.
- US Bureau of Labor Statistics. Occupational employment and wage statistics for the warehousing industry. Primary. Occupational wages including metropolitan estimates.
- US Bureau of Labor Statistics. Geographic guide to labor statistics programs. Primary. Explains which programs publish at which geographic level.
- Federal Energy Regulatory Commission. Explainer on the interconnection final rule. Primary regulator source on the move to cluster studies.
- Lawrence Berkeley National Laboratory. Queued Up, interconnection queue analysis. Independent national laboratory. Source of the queue duration and volume figures cited.
- US Census Bureau. Value of construction put in place, including private warehouse construction. Primary. Independent construction spending data, which is spending rather than occupancy.
- US Census Bureau. Housing Vacancy Survey. Primary. Confirms that the federal vacancy series covers residential property only.
- American Planning Association. Summary of peer-reviewed research on warehouse zoning actions. Independent professional planning body summarizing research published in its journal.
- CoStar Group. Regulatory filing describing product use for vacancy, absorption, and rental rate calculation. Interested source: a for-profit data vendor selling subscriptions. Cited to establish the provenance of market metrics.
