When a sortation line goes down and the next peak is eight weeks out, the decision usually lands between three paths: a custom-engineered cross belt sorter, a modular quick-ship platform, or a refurbished system. I've coordinated 60+ rush deployments over nine years as a project lead at a material handling integrator, and in my experience the sticker price is almost never where this decision lives. Lead time is. Rework risk is. Re-commissioning cost is.
Here's how the three stack up across the dimensions that actually decide a rush project. Same framework I use when I'm triaging an order that came in at 6 a.m. with a customer asking whether we can hit an install date before their busy season.
The comparison framework
(A) Custom-engineered cross belt sorter, (B) modular quick-ship platform — think the E24-style programs a lot of OEMs run now — and (C) refurbished or used equipment. Four dimensions: lead time, upfront cost, total cost of ownership, and reconfiguration flexibility. Anything else is a detail.
Dimension 1: Lead time
This is the whole job, basically. If you can't ship on the date the client needs, the other dimensions don't matter.
- Custom build: 16–24 weeks from design sign-off to first sorted carton. Some specialty configurations push past 30 weeks.
- Modular quick-ship: 4–8 weeks for standardized footprints. A distributor with pre-positioned inventory can compress that further, but you're color-by-numbers on layout and carrier count.
- Used/refurbished: 6–12 weeks once you count rigging, trucking, electrical, controls retrofit, and re-commissioning. "Available now" is the sales pitch; the reality is measured from the day the unit is powered up, not the day it's quoted.
So the counterintuitive part: used equipment looks fastest because the machine already exists. In practice, for anything above a very simple sortation config, re-commissioning eats more of the calendar than a modular platform does.
Dimension 2: Upfront cost
Rough ranges from quotes I've handled and publicly discussed industry figures in 2024–2025 — treat these as ballparks, not lists, because every quote is site-specific:
- Custom mid-size sorter (roughly 6,000–12,000 items/hour, 60–120 carriers): $350,000–$850,000 installed. Specialty builds go significantly higher.
- Modular quick-ship in the same throughput envelope: often 10–25% less than a comparable custom line, because the OEM isn't engineering from scratch.
- Refurbished: typically 30–55% of new sticker. That's the number that gets everyone's attention.
Then again, sticker is where the comparison stops being useful. Which brings us to the dimension that cost-focused buyers underweight.
Dimension 3: Total cost of ownership
They warned me about re-commissioning costs on used systems. I didn't listen once. We ate roughly $40,000 over budget on a 'deal' refurbished unit we placed for a parcel client in 2023 — new controls, new photo-eyes, a carrier belt rebuild that looked fine on the inspection report, and two weeks of engineering labor nobody line-itemed. The $55,000 price tag became an $95,000 project. Nobody at the client was angry about the cap-ex number. They were angry about the two weeks.
Put another way: for a modular platform, the OEM's predictability is the product. You're paying a small premium for a parts book that actually matches the machine and a service network that already knows your controller.
I can only speak to mid-size e-commerce and parcel profiles. If you're running cold-chain or grocery, washdown ratings and duty cycle requirements shift the calculus hard — the used market for those applications is thin, and the re-engineering risk goes up enough that I usually push clients back toward new.
Dimension 4: Reconfiguration flexibility
Custom gives you the exact footprint you asked for and almost nothing more. Adding carriers or extending the loop later usually means a second engineering cycle.
Modular platforms are built around the idea that you'll bolt on later — but only within the OEM's module catalog. If your six-month projection involves throughput jumping by 40%, you need to model that before you buy, not after.
Used is the least flexible of the three, and that's fine, as long as you've accepted that the configuration you're buying is the configuration you keep.
What the money actually buys
Here's the honest summary. You're not really choosing between price points — you're choosing between a time guarantee, a spec guarantee, and a flexibility guarantee. You can usually have two. Rarely all three.
So glad we pushed one client toward a modular platform last spring. We almost went custom to save roughly $60,000 on paper against a neighboring quote, and we would've missed their Memorial Day ramp by three weeks. Cost the client a fraction of what that lost weekend would have.
Which one fits your situation
If the deadline is the hard constraint — peak season, a contract penalty clause, or a customer already buying from someone else if you miss — the modular quick-ship path is almost always the right call. You're paying a modest premium for schedule certainty, and schedule certainty is worth more than the premium in almost every rush project I've run.
If your requirement sits outside the standard envelope — unusual footprint, product mix that breaks the module catalog, or throughput the platform can't reach — custom is the answer, but only if you have 16+ weeks. Do not start a custom engineering cycle against an 8-week deadline. I've seen it tried. It ends badly for everyone.
If you have 12+ weeks and in-house engineering or a strong integrator relationship, refurbished can genuinely work. Used isn't the enemy — bad used is the enemy. A used unit with complete documentation, a clean controls audit, and a service path for the critical components is a different animal than the bargain-bin refurb that shows up on a skid with a 'good luck' invoice.
Bottom line: match the path to the deadline first, the requirement second, and the budget third. In rush work, the sequence is basically the whole job. The cost of getting the sequence wrong shows up about two weeks before go-live, and by then there's no re-quote that saves you.