← Percision · Blog

Build, Buy, Partner, or Walk Away: An NPV/IRR Scenario Model for Logistics & Supply Chain Decisions

Direct answer: For any major logistics capability decision—warehouse automation, a TMS/WMS platform, a last-mile fleet, or a new lane—the right path (build, buy, partner, or walk away) is the option with the highest risk-adjusted net present value (NPV) and an internal rate of return (IRR) that clears your cost of capital across a realistic range of demand and cost scenarios. Don't decide on gut feel or vendor pitch; model each path over the same time horizon, run best/base/worst cases, and pick the one that wins in the base case and survives the worst.

Logistics decisions are unusually well-suited to NPV/IRR modeling because the cash flows are concrete: capex, labor, fuel, throughput, and contract terms are all measurable. The trap is that "build" and "buy" often look similar on a single-point estimate and diverge sharply once you stress-test volume, fuel prices, and utilization.

Why NPV/IRR Scenario Modeling Fits This Decision

A build-vs-buy-vs-partner call has four candidate cash flow streams, and each behaves differently under uncertainty:

NPV discounts each stream's cash flows to today's dollars so you compare them on equal footing. IRR tells you the return each path generates against your hurdle rate. Scenario modeling forces you to admit that logistics cash flows swing with variables you don't control—diesel, e-commerce demand, labor markets, lease rates.

A Concrete Walkthrough

Say a mid-market distributor is deciding how to handle fulfillment for a growing region.

Step 1 — Define the decision and horizon. Pick a common evaluation window (7–10 years is typical for warehouse and fleet assets, matching depreciation and lease terms). All four options get measured over the same period.

Step 2 — Build the cash flow lines for each path.

Step 3 — Set the discount rate. Use your weighted average cost of capital (WACC). Logistics carries operational and demand risk, so the hurdle rate should reflect it. If your WACC is, say, in the low double digits, an option's IRR must exceed that to create value.

Step 4 — Run three scenarios per path.

Step 5 — Read the results. Good looks like this:

The insight usually isn't which number is biggest—it's which path is robust. Partnering often trades some base-case NPV for a much better worst case. Building wins only if you're confident in volume. Walking away wins more often than pride allows.

How Percision Helps—and When a Spreadsheet or Consultant Is Enough

Disclosure: I work on content for Percision (percision.app), so treat this as one option, not the only one.

Percision is a strategic intelligence platform that runs your business context through structured reasoning steps across specialist models to produce board-ready output—DCF valuations, scenario analyses, and Excel-exportable models with audit trails—in minutes rather than the weeks a manual build takes. For a build/buy/partner/walk-away call, it can generate the four cash flow scenarios, discount them at your WACC, surface the breakeven variables, and hand you a board deck plus an editable model you can adjust. It's positioned as a co-pilot, not an autopilot: your team supplies the assumptions and owns the decision. (Independent research from sources like BCG and Harvard Business School has found AI tools can meaningfully speed up knowledge work on structured tasks—useful context, not a guarantee about any single decision.)

When you don't need it: If you have a clean two-option comparison and a competent FP&A analyst, a well-built spreadsheet is entirely sufficient—and cheaper. When to hire a human consultant instead: if the decision hinges on soft factors a model can't capture—labor relations, regulatory exposure, a distressed acquisition target's hidden liabilities—senior operator judgment matters more than speed. Use Percision to compress the analytical grind and pressure-test scenarios; use people for the judgment calls at the edges.

The right sequence: model first, decide with humans, then build the execution plan against the winning scenario's assumptions—so you know exactly which variable to watch after you commit.

What this looks like when the analysis is actually run

Lease, buy, or reprice what you already have — three options with very different reversibility, priced side by side.

The subject is Ridgeway Freight Systems, a sample company profile we use for testing rather than a customer: a regional LTL carrier, $284M revenue, 18 terminals, 620 drivers.

Excerpt from a real Percision run · Competitive Positioning (T9) · sample company profile

Lease. 50–75 reefer trailers under a 36-month master lease: $3–5M over 36 months — $1.0–1.5M of Year 1 deposits, $1.2–1.8M of Year 2 lease payments, $0.8–1.2M of Year 3 maintenance and wash facilities. Base case 28–36% IRR on $4M; payback 22–26 months.

Buy. $8–12M for outright reefer purchase — roughly two and a half times the lease commitment for the same capacity.

Reprice what exists. 4–6% rate increases across 14 contracts, 61% of the $82M dedicated book, on a $0.8–1.2M annual retention pool, for a $3.3M annual operating-income lift at 275–413% ROI and zero incremental fixed costs.

What the lease buys that the purchase does not. Terminate if utilization is below 65% for two consecutive quarters or the reefer segment operating ratio exceeds 96.0 for six months; redeploy tractors to dry-van dedicated and return the reefers to the lessor at Month 24 with no penalty.

The ramp each option is judged against. Year 1: $2–4M from 2–3 pilot contracts, 50 reefers at 60% utilization. Year 2: $6–9M from 5–7 contracts, 65 reefers at 70%. Year 3: $12–18M from 8–12 contracts, 75 reefers at 75%, at an average contract value of $4.2M and an 8% operating margin after lease and wash-facility costs. Pilot contracts signed by Month 9.

Go / no-go gates before the next phase is funded
PhaseGate metricTargetDeadline
Foundation (0-6 months)Master lease signed AND 2 pilot customer RFPs issuedLease executed; RFPs live by Month 6Month 6
Traction (6-18 months)Revenue run-rate ≥$6M AND turnover ≤50% on reefer fleet$6M annualized; turnover ≤50%Month 18
Scale (18-36 months)Operating income contribution ≥$1.0M AND portfolio dedicated concentration ≤45%$1.0M OI; concentration ≤45%Month 36

The IRR comparison is not what decides it. 28–36% on a lease is a decent return and 275–413% on repricing is a much better one — but they are not substitutes, because only the reefer entry reduces a top-two customer concentration of 61%.

The lease-versus-buy line is where the model earns its keep. Paying $3–5M instead of $8–12M costs some economics over three years and buys a Month 24 exit with no penalty. For a carrier with a 2027 put obligation and deferred fleet capex, optionality is worth more than the spread.

Read a complete Percision report — every page, no email required.

FAQ

What discount rate should we use for a logistics build-vs-buy model? Use your company's WACC as the floor. If the specific project is riskier than your average operations (a new geography, unproven automation), add a risk premium. The option's IRR must exceed this rate to create value.

Why include "walk away" as a formal option? Because the status quo has real cash flows and real risks, and it's the honest baseline every other path must beat. Modeling it prevents you from approving a marginal build just because someone already sunk effort into planning it.

Can NPV alone decide this? No. Pair NPV (dollar value created) with IRR (return vs. hurdle rate) and scenario stress-testing (robustness). A high base-case NPV that collapses in the worst case is a worse choice than a smaller, sturdier one.


Want to run all four options—build, buy, partner, walk away—through a scenario model and get a board-ready deck plus an editable financial model? Try Percision and keep your leadership team in control of the call.

Ready to run this on your company?
A free Percision diagnostic turns the analysis into a decision with owners and numbers — one click from this article.
Run the free diagnostic →
Get the full State of AI Strategy 2026 report
The research, the method, and the pre-registered tests — plus occasional notes on governed AI strategy. No spam; unsubscribe anytime.