Choosing where to open a new warehouse, branch, or distribution center — balancing transport cost, service time, and customer coverage (known in the literature as Facility Location).
In plain words
Sound familiar?
- The company has grown, order volume is 2–3× higher; from the current warehouse, shipping costs across the country have ballooned, but the 'where do we open the new warehouse' decision keeps getting deferred
- Customers want 'same-day' or '24-hour' delivery; the current location can't meet that promise, and which region to invest in is unclear
- The investment question 'should we open in city A or city B' has been sitting open on the leadership table for months
- Site feasibility is done on intuition — 'next to the factory' or 'near the port' — with no mathematical analysis
- Current warehouse capacity is exhausted, but whether to expand it or open a new one is unsettled
- In the dealer/branch network, 'which branch to close and which to expand' has been pending for years
- There are acquaintance companies that realized 6–12 months after the investment that the decision was wrong — and the reasons aren't clear
Why it matters
How it's solved
Technical depth
How it's solved
Technical depthOne-liner: Open the new warehouse not near the factory but near the center of demand — the geographic average of your customer order volume. Then nudge that point 30–50 km to honor rent and service-time (24-hour / 48-hour reach) constraints. ‘Put it next to the factory’ usually misses optimum by 15–30%.
What the software is really doing is this: instead of comparing two or three sites by gut, it mathematically evaluates all candidate locations, balancing transport cost, service time, rent, taxes, and capacity constraints together. Three stages:
1. It builds the demand map. Customer locations (by city/district), each customer’s annual order volume, order frequency, delivery sensitivity (hours/days), and unit weight/volume. Existing warehouses and distribution centers, current routes, transport tariffs (distance × volume). The data flows from the order system automatically, or is entered once into a clean table.
2. It scores candidate sites. The software evaluates every possible location — city centers, industrial parks, organized industrial zones, customs-adjacent sites, ports. For each candidate it computes numerically: annual total transport cost to all customers, service-time profile (how many customers within 24 hours, within 48 hours), estimated rent/tax/labor cost, capacity ceiling. Using operations research (a discipline that uses math and computing to solve business-decision problems) algorithms — p-median (where to place N facilities so average distance to customers is minimum) and capacitated facility location — it picks the optimal combination from the candidate set.
3. It supports decision-making through scenarios. The software produces several scenarios: ‘single central warehouse’ / ’two regional warehouses’ / ’three local warehouses’ comparison. For each scenario, a 5-year total cost of ownership (TCO), service level, and sensitivity analysis (what if demand grows 20%, what if rent shifts 30%). The leader’s decision desk ends up with four or five durable numbers.
It does not replace the manager’s judgement; think of it as a calculator that does what you do with 5 scenarios, but with 500 scenarios and built-in sensitivity analysis. The decision is still yours, but the investment case is numerical.
Alternatives
Gut feel + one or two scenario comparisons
FreeFree
Who it fits: 1 facility investment, low customer variety, simple distribution
- + Zero cost
- + Fast decision (days, not weeks)
- + Leadership stays close to the field
- − Not every possible site is considered
- − Service level can't be quantified
- − 5–10 year operational cost outlook is missing
- − No sensitivity analysis — what if demand grows 20%?
One-off consulting study
Enterprise$15,000–60,000 project-based (regional SMB pricing)
Who it fits: 1–2 large facility investments, 3–6 month decision window
- + Professional analysis, experienced team
- + Numerical justification for the investment
- + Local knowledge (taxes, incentives, industrial zones)
- − Project takes 2–4 months; a fast-changing market may move on
- − One-off — re-engagement needed when demand shifts
- − Decision logic stays with the consultant, not in-house
Local network design software
Enterprise$2,500–8,000/year project-based or subscription (regional SMB pricing)
Who it fits: 3–10 existing facilities, periodic re-evaluation, existing order data set
- + Local-language interface and support
- + Local map and distance matrix integrated
- + On-screen scenario comparison
- − Optimization engine is usually basic — limited on large candidate sets
- − Sensitivity analysis is limited
- − Stochastic demand (demand uncertainty) model is weak
International specialized network design software
Enterprise$50,000–300,000/year licence, or project $100,000–500,000
Who it fits: 5+ facilities, multi-country, complex tax and incentive structures
- + Mature: multi-echelon supply chain, stochastic demand, multi-criteria optimization
- + Tax and incentive libraries kept current
- + Algorithms hardened over years
- − High licence and consulting cost
- − Data preparation takes 2–4 months
- − Local-context constraints (industrial zones, regional incentives) need custom configuration
Custom build on an open-source solver
Open SourceLicence free; 8–16 weeks of internal development, or $60,000–250,000 of consulting
Who it fits: Large distributor with an in-house data and OR team
- + No licence cost
- + Fully customizable to your cost and service structure
- + Ongoing decision-support instead of a one-off study
- − Requires internal technical capacity
- − Data quality issues feed misleading conclusions
- − Best positioned as a continuous decision-support system, not a single analysis
Recommendation
Ask in the meeting
- How is the candidate site set built — only from your input, or does the system have a full database of cities, districts, and industrial zones ready?
- Is the transport cost based on real tariffs (km × weight) or simple Euclidean distance? Is there a library of regional carrier tariffs?
- Are multi-facility scenarios (e.g. 1 vs 2 vs 3 warehouses) and different capacity options compared automatically?
- How detailed is the sensitivity analysis (demand +20%, rent +30%)? Is Monte Carlo simulation supported?
- Are local tax, incentive zones, and rent ranges kept up to date in the library?
- Is the 5- and 10-year total cost of ownership (TCO) calculation standardized? Which assumptions are visible?
- How do you structure the pilot — how many scenarios, how many weeks, what is the success bar?
- If we stop working with you, how do we get our scenario models, demand data, and cost library back? Is there a standard export format?
Technical details
Editor’s note
On the leadership desk this problem is called “where to put the warehouse”, “where to open the new branch”, or “the site decision”. The academic name is the Facility Location Problem. It began with Alfred Weber’s 1909 theory of the location of industries, matured in the 1960s with mathematical programming, and is studied today with extensions covering stochastic demand, multi-echelon supply chains, and carbon footprint. Without that vocabulary, in a consulting or software demo you cannot tell whether the “location analysis” being offered is real optimization or just visualization on a map.
The point most often overlooked in this segment: many consultancies and tools advertise “location analysis” but underneath they only compute the center of gravity — the geographic mid-point weighted by customer demand. That’s a fine first pass for a single facility; but for multi-facility, capacity-constrained, or tax-incentive-sensitive decisions, that math is incomplete. In any demo, insist on running 3 scenarios (1 vs 2 vs 3 warehouses) with different capacity options and 5-year demand growth assumptions, and ask ‘how is total cost of ownership computed?’
A step-by-step path for an SMB
Stage 1 — Measure first, plan later. For at least 12 months, log four things:
- Customer locations (by city/district, ideally coordinates) and annual order volume
- Transport cost from the current warehouse to each region (km × weight × tariff)
- Current service time (hours or days from order to delivery)
- Current warehouse operating cost (rent, labor, equipment, taxes)
Without this baseline, neither a vendor nor a consultant can give a defensible recommendation.
Stage 2 — Build the candidate-site list. Compile 15–30 city/district or industrial-zone candidates. For each: estimated rent range, labor cost, incentive status, transport infrastructure (highway, port, airport proximity). This list is your knowledge capital — any serious vendor or consultant will ask for it first.
Stage 3 — Modeling phase. A 6–16 week analysis period, depending on investment scale. Define the success criterion in writing, before the analysis: e.g. “total distribution cost down 10% within 18 months of investment, customer service time under 12 hours.”
Stage 4 — Decision and rollout. A 6–12 month facility setup period after the decision. During rollout the current and new warehouses operate in parallel; over 3–6 months the order volume migrates to the new facility in stages.
Risks — what can go wrong
- Demand forecast error. The investment is for 5–10 years, but the market structure can shift in that time. Stochastic demand scenarios (demand +20%, demand –20%) must be computed.
- Focusing on a single number. Picking the city with the lowest rent ignores total cost — rent is typically 15–25% of TCO, labor 40–50%, transport 20–30%.
- Local regulation and incentives. Industrial zones, tax breaks, and regional incentive areas can shift the decision by 2–3 percentage points. Local regulatory advice is essential.
- Vendor lock-in. Software or consultants that keep the decision model and scenario assumptions in proprietary format make follow-up hard. Put a clause in the contract: “We can export our model and data in standard open formats (CSV, JSON, or similar) at any time, on request.”
Related cautionary lesson (will be linked once published): “A mid-sized distributor placed its second warehouse in city B instead of city A, reversed the decision after 14 months — what they missed.”
A technical view of the solution method
This section holds what you’ll need when talking to a consultant or a software team. The investment decision lands as 4–5 numbers on the leadership desk, but behind those numbers are the methods below.
The main approaches used for facility location:
| Approach | Typical use | Data need | Decision logic |
|---|---|---|---|
| Center of Gravity | Single facility, first pass | Low | Demand-weighted geographic mid-point |
| p-Median | Pick a fixed number of facilities | Medium | Minimize weighted distance from customer to nearest facility |
| p-Center | Service-time-critical | Medium | Minimize farthest customer distance |
| Set Covering | Coverage within a fixed distance | Medium | Every customer within X km of at least one facility |
| Capacitated Facility Location (CFLP) | Real constraints | High | Optimum site + capacity + assignment |
| Stochastic MIP | High demand uncertainty | High | Expected cost across demand scenarios |
In practice: single facility with quick decision → center of gravity. 2–5 facilities with capacity constraints → CFLP. Multi-year with stochastic demand → stochastic MIP or scenario-based analysis.
Objective function choice changes the shape of the solution:
- Total transport cost: “Minimize freight” — fits freight-heavy operations
- Service level (e.g. % delivered within 24 hours): “Customer quality” — fits premium-service operations
- Total cost of ownership (TCO): “5–10 year economics” — fits long-term investment decisions
- Carbon footprint: “Sustainability target” — fits operations under customer sustainability pressure
Most real deployments use a weighted blend of all four — most commonly TCO × service level.
Academic references
Listed in the sources block of this page’s frontmatter. Facility location has been one of operations research’s oldest fields since the 1960s; current work extends it with sustainability, multi-echelon supply chain, and omni-channel retail dimensions. INFORMS Interfaces and the European Journal of Operational Research archive carry deployment case studies tied to real distribution operations.
Sources
- Daskin, M. S. (2013). Network and Discrete Location: Models, Algorithms, and Applications (2nd ed.). Wiley. The standard textbook on facility location.
- Melo, M. T., Nickel, S. and Saldanha-da-Gama, F. (2009). Facility location and supply chain management — A review. European Journal of Operational Research, Vol. 196 — modern review of the field.
- Drezner, Z. and Hamacher, H. W. (eds., 2002). Facility Location: Applications and Theory. Springer. Classic and applied review of facility location.
- INFORMS Interfaces — case studies of facility location deployments in distribution and retail. informs.org/Publications/Interfaces
Glossary
- Facility Location
- Choosing where to place a new facility (warehouse, plant, branch, hospital) — mathematical optimization across demand points and costs.
- Network Design
- Designing the physical and flow structure of a supply or distribution network at the strategic level — a long-term investment decision.
- MIP
- An optimization model where some decision variables are forced to be whole numbers (e.g. number of trucks, number of shifts).
Related problems
From One Node to Another — How Do I Compute the Shortest Path on a Weighted Graph?
For SMBs that need to compute the fastest or shortest route between two points: 10-50-vehicle field-service teams (plumbing, electrical, appliance repair), urban courier/parcel operations, or dispatch centres coordinating emergency response. Every day brings hundreds of 'how do I get from A to B fastest right now' questions; the answer shifts with traffic, road closures and vehicle type. A wrong route costs the technician one or two jobs missed for the day, the courier a late delivery, and the firm a customer. Manual or by-eye routing typically leaves 20-60 wasted minutes per vehicle per day on the table compared with a network-aware route calculation.
Multiple Plants, Multiple Customers — How Much Does Each Plant Ship to Each Customer to Minimise Total Freight?
For food, packaging or textile producers shipping weekly from 3-8 plants or regional warehouses to 20-100 customers. The weekly decision is: which plant ships how much to which customer, given fixed plant capacities, stated customer demands, and a different per-unit cost (distance + vehicle + contract terms) for each plant-customer pair. The goal is the lowest total freight bill across the network. A 'nearest plant' or 'we've always done it this way' habit typically leaves 10-20% extra fuel and vehicle cost on the table compared with a systematic allocation.
One Vehicle, Many Stops — In What Order Should I Visit Them All to Minimize Total Distance?
You run a field technician visiting 8-15 customers a day (HVAC, lift servicing, white-goods repair), a single-vehicle supplier tour by a sales rep, or a PCB drilling machine sequencing 500-5,000 holes. All of them face the same core call: given N points, in what order should a single vehicle or head visit each one and return to the start. Get the order wrong and a field service vehicle burns 80-200 TRY/day extra in fuel and driver hours, a PCB line takes 15-30% longer per part, and the last customer of the day misses their delivery window. At 50 stops, a hand-built sequence runs 20-40% above the true minimum; as the number of stops grows, the gap from intuitive ordering compounds.