Look — the biggest uncontrolled line in your FY28 transportation budget isn't the rate increase. It's the cost of being slow to change. Published GRIs are at least a number you can argue with; the realized increase runs 100 to 200 basis points above the headline (S2 — the surcharge machinery, well documented), and you can audit that too. But the weeks your stack needed to encode a new fee, certify a new carrier service, or open a new geography? Nobody budgets that. It shows up anyway — as margin you didn't capture and exposure you couldn't shed — and it repeats every year until you price it.
This one's for the CFO running the mid-year review. Two passes: what shipping actually cost since January, and how fast the machine underneath it can turn.
Pass one: decompose the half-year
Same discipline as January, now with six months of evidence. Pull the invoice file and split your effective rate movement into base, fuel, DIM, and accessorial — the Surcharge Intelligence Index decomposition. The output is one exhibit: your effective increase versus the published one, and which surcharge families drove the gap. That sets the FY28 freight-inflation assumption on evidence instead of press release, and it sets the surcharge-drift reserve — because if your network has been running a point or two over published, FY28 deserves a budget line that says so out loud instead of an explanation in Q3.
While you're there, add the line almost every Midwest-exposed budget is missing: weather contingency as a number, not a shrug. If your network runs through this region — and with six Class I railroads meeting in one metro, most national networks do — last winter's diversions, expedites, and lake-effect re-routes are sitting in your H1 invoices, priced. Total them. That's the contingency line, evidence-based, and the gateway's dwell behavior at the Elwood and Joliet ramps this spring is your early read on how hard next winter's version bites. [S-cite: H1 gateway dwell trend vs. prior-year Q4 service performance].
Pass two: score time-to-change
Now the pass nobody runs. List every change your shipping stack absorbed in the last twelve months — new surcharges encoded, carrier services added, rate-card updates, a geography or channel opened. For each: calendar days from decision to in-production. That distribution is your stack's metabolism, and it converts to dollars directly. A demand fee encoded in two days is rated correctly from day one; the same fee encoded in six weeks is six weeks of mispriced freight. A regional carrier certified in three weeks captures a contract-season opportunity; in five months it misses the season entirely. Slow stacks pay rate increases twice — once on the invoice, once in the lag.
This is where the developer-first vendors earned their position, and the credit is factual: EasyPost shipped the first dev-first multicarrier API back in 2012 — the "Stripe for shipping" pitch, Y Combinator S13 (S16 — the YC writeup) — and Shippo built a real business on the same architectural bet. An API-first stack genuinely moves the time-to-change needle for parcel. Two qualifiers before you write that check. First, grade API quality the boring way: carrier certifications shipped per quarter, documentation freshness, sandbox fidelity to production — certification cadence is a real, measurable discipline, the kind sustained over years (ProShip's twelve consecutive years of FedEx Diamond-level compatibility shows what that looks like as a track record, S18 — their announcement). Demos don't certify carriers; programs do. Second, a parcel-only API doesn't model the seam that sets your costs in this region — the intermodal handoff where ramp dwell becomes injection timing becomes parcel spend. Fast parcel rating downstream of an unmodeled rail delay is a fast wrong answer.
Score it like a distribution, not an average, same logic as latency: the median change probably looks fine, and the slow tail is where the money went. One six-month carrier onboarding buried among ten two-week fee updates is the line item worth the meeting.
The FY28 budget, restructured
Four lines that earn their place:
- Freight inflation — your decomposed effective rate, not the headline.
- Surcharge-drift reserve — sized from your own H1 gap to published.
- Weather contingency — last winter's diversion costs, totaled and indexed.
- Time-to-change investment — whatever closes the worst gaps your pass-two audit found, priced against the mispriced-freight lag it eliminates.
The tradeoff is honest: pass two takes real engineering time to instrument, and the payoff lands in avoided cost, which CFOs rightly distrust. So anchor it in last year's actuals — the lag costs already incurred, at invoice level — not in projections. Slow already cost you money. The audit just finds the receipts.
Concrete ask: send us your H1 invoice file and your change log and we'll run both passes — the Surcharge Intelligence Index decomposition and the time-to-change scorecard — and hand back the four budget lines with the evidence attached. Two weeks, in time for budget season. The increase you can negotiate. The slow you have to fix. Budget both and FY28 stops surprising you.