How to compare ASRS quotes when no two vendors quote the same thing
Six proposals came back. One is a unit-load crane system, one is shuttles, one is a fleet of robots. Which one’s cheapest? You can’t tell yet.
I wrote a short post on this in June. It’s a problem a lot of operations teams have lived through: sitting in a conference room, staring at a spreadsheet where the numbers refuse to line up. So here’s the longer version.
The core problem is simple. Every vendor proposes the system they make, sized with their own throughput math, and draws the scope line wherever suits them. The cheap quote might be cheap because it left out the hard parts. The expensive one might be gold-plated, or it might be the only one that actually solves your problem. The price means nothing until all of them sit on the same baseline.
Start before the RFP goes out
Normalizing is far easier if the vendors all quote against the same basis of design. That means your order profile, design-day and peak throughput, SKU counts and dimensions, storage capacity with a growth assumption, operating hours and the building. Put it in the RFP as data, not adjectives. “High throughput” invites six interpretations. “1,850 order lines an hour at peak, 62% single-line” gets you six answers you can check.
Ask every vendor to return pricing in the same structure, too. A response template with fixed line items saves you weeks later.
Find what’s missing
This is where most of the money hides. In my experience the items most often excluded, assumed by the customer or buried in a footnote are some version of these:
- Fire protection in or around dense storage, which can be a large number on its own
- Slab, floor flatness, structural and seismic work, plus engineering stamps and permits
- Power drops, network, IT hardware and the interface to your WMS or ERP
- Software licensing and annual support, spare parts packages, training and post-go-live support
Build a matrix with every scope item down the side and every vendor across the top. Mark each cell included, excluded or unclear. Then send the unclear ones back as written questions. You’ll learn a lot from how fast and how specifically each vendor answers.
For anything excluded, put in your own estimate so each proposal now covers the same scope. That adjusted number, not the quoted one, is what you compare.
Check the throughput, not just the price
Two systems at the same price aren’t the same deal if one only hits your rate on an average day. Ask each vendor to show the math: cycle times, the order profile they assumed, and the utilization at peak. Run your worst week through it. A system designed to run at 95% utilization on paper has no room for a bad day, a new customer or a vehicle out for maintenance.
Ask what happens when one component fails. One crane down in a single-crane aisle means that aisle is dark. One shuttle or robot down in a fleet usually means a slower hour. That difference belongs in the comparison.
Score on total cost of ownership, with weights you agree on first
Capital cost is one line. Over ten years you’ll also pay for software support, service contracts, spare parts, energy, the labor the system still needs, and probably an expansion. Put all of it on one sheet per vendor.
Then score the things money doesn’t capture: technical fit, delivery and execution risk, US service footprint, the vendor’s financial strength and, now, regulatory risk on imported mobile equipment. On a recent selection I ran, six vendors were scored this way with weights the client’s team agreed before a single proposal was opened. Agreeing the weights first matters more than any specific number. Otherwise the scoring quietly bends toward whoever the loudest person in the room already likes.
Who should do the normalizing
Someone with nothing to sell you. Integrators and manufacturers do a lot of good engineering, but none of them can neutrally score a competitor’s proposal. That’s the main reason owner’s representative work exists, and it’s usually the highest-value hour of the whole project.
If you’ve run an automation RFP, I’d like to hear how you handled proposals that weren’t really comparable. The numbers rarely line up on their own.
More insights
The FCC robot rule, from the buyer's side
What the July 2026 Covered List update means for open quotes, installed fleets and the conversation with your CFO.
The first 180 days of selling automation in North America
For manufacturers from Asia-Pacific and Europe: what to build, who to hire and which mistakes cost a year.
Shuttle buffer vs. more conveyor: when the buffer wins
More conveyor feels cheaper. It often isn't, especially when the sequence matters as much as the space.