"We're bursting at the seams."
My operations VP stopped by my desk on a Tuesday morning in September 2023. We'd just closed a new customer contract that was going to push 30% more volume through our Ohio warehouse, and the team was already stacking product in the aisles.
Quick context on me: I'm the admin buyer for a roughly 400-person company with three warehouse sites (Ohio, Texas, Pennsylvania). I handle procurement for office supplies, facilities services, and material handling equipment—about eight vendor relationships, $1.2M a year, give or take a few hundred thousand. Storage equipment isn't my home turf. The warehouse team usually runs that. But procurement runs through me, so this landed on my plate.
I figured, how hard can this be? We'll buy more racking.
Starting with the obvious: wholesale racking
My first call went to our wholesale warehouse racking supplier. Quote, unit price comparison, standard process. They came out, measured the bay spacing, and gave me a number for additional pallet positions.
The per-position price looked fine—around $180-220 installed.
But the most useful thing I did was ask them, directly, how much would actually fit. The answer was uncomfortable. Our Ohio warehouse floor was already maxed out. We could bolt more pallet positions onto the existing vertical height—we've got 22 feet to work with there—but standard racking math meant tighter aisles and slower pick speeds. Basically, buying more wholesale racking wasn't going to solve the problem. We needed more storage in the same footprint.
That's what got me looking at automation.
The vertical lift module phase (and what I got wrong)
The warehouse manager asked, almost as a joke: "What about those automated towers?"
I knew roughly what she meant. We'd seen a few vertical lift module units at a trade show. The logic made sense: vertical storage, modular units, small footprint.
So I reached out to a few vertical lift module suppliers—three, I think. Maybe four, I'd have to check my inbox. Each one walked the floor and sent back pricing.
This is when I actually had to think hard about what to look for in a vertical lift module supplier. My first pass at a checklist was embarrassingly short: price, lead time, warranty. That was it.
The list that actually mattered looked more like: single-cycle pick time, multi-user access capability, ceiling height requirements, integration with our existing WMS, service response coverage, and—most critically—whether they could provide deployed references at companies our size and in our industry.
That last item eliminated two suppliers. One had plenty of cases but they were all single-site, large-footprint operations. We needed a layout that could work across three smaller facilities. Very different problem.
Vertical lift modules didn't win the project—at least not standalone. At our throughput, we'd have needed to buy unit-by-unit, and the capital cost to replicate across three facilities wasn't going to fly.
The AutoStore ASRS conversation and a mental shift
Then an integrator brought up AutoStore ASRS—specifically, the cube storage approach. Bins stacked ten-plus levels, robots riding the grid on top.
I was resistant at first.
The "automation means a full facility rebuild and six months of downtime" idea was a leftover from about 2017 in my head—back when cube storage was custom, single-site, and risky. Today that's not the reality. Modular cube systems let you start with a few thousand bins per week and scale up. That misconception cost me a couple of weeks.
Once I actually dug into the AutoStore storage system mechanics, the logic held up. The aluminum grid, the bins, the robots—it's a dense storage play. Throughput scales roughly linearly with robot count. Expansion means adding grid sections or robots, not building a new wing.
But there are real limits. Bin dimensions cap what you can store—you're not putting a four-foot item in one. Maintenance access needs to stay clean. And the WMS integration work on the front end is not trivial.
So we didn't end up replacing everything with cubes. The final design was a hybrid: fast-movers in the automated cube, long-tail inventory on conventional pallet racking. More on that in a second.
Picking a storage system supplier (and the one hard rule)
We engaged about six storage system suppliers total. We used the AutoStore ASRS configuration as our cost and throughput benchmark for the automated side.
I mentioned earlier that I'd been burned before on procurement. Early in my career I ordered from a vendor who couldn't produce a proper invoice—handwritten receipt only. Finance rejected the expense report and I ate $2,400 out of my department budget. Now invoicing and documentation capability is the first filter I apply. This time, I made it a hard spec requirement.
The other hard requirement I wrote into the spec was written throughput calculations. I told every supplier: don't give me a theoretical throughput number. Show me the assumptions behind it. SKU distribution, order profile, peak vs. average volume. Put it in writing.
There's a relevant standard here, though I didn't expect to use it: FTC advertising guidance requires that marketing claims be truthful, not misleading, and substantiated with evidence (ftc.gov). I didn't expect suppliers to read the FTC guides, but I applied the same standard. If a throughput claim couldn't be backed by a documented model, it didn't go into the contract.
One supplier quoted a number that looked impressive until we asked for the assumptions. They sent a one-page PDF with roughly six bullet points and no SKU breakdown. That was a red flag. We passed.
The outcome
We signed in early 2024. Budget was approved at roughly $800K; final landed around $650K, because the first AutoStore cube we deployed was smaller than the original design.
Looking back, I should have built a formal evaluation matrix before any quotes came in. My first round of "comparison" was just a stack of documents—one PDF, one Excel sheet, one email—incompatible formats, impossible to compare cleanly. If I could redo that decision, I'd insist on a common set of fields for every supplier response. But given what I knew then—which was basically nothing about automated storage—that mistake was probably inevitable.
One more thing I only saw clearly after the fact: one of our three facilities didn't need automation at all. Two got the cube system; the smallest site just needed re-racking and a layout change. That "don't automate" call came from analyzing their actual order profile rather than the supplier's demo scenario.
What I'd tell the next admin buyer doing this
If you're staring down a storage capacity problem and someone just told you to buy more pallet racking, here's what I'd share from this project:
- Modular beats all-at-once. Cube storage can start small. Don't let the "automation = six-month shutdown" thinking—true maybe five years ago—stop you from at least pricing it.
- Put throughput assumptions in the spec. Every quote should document SKU profile, order profile, and ramp-up plan. No exceptions.
- Vertical lift modules and cube storage solve different problems. VLM units make sense for high-turnover, small-item, narrow-footprint storage. Cube storage makes sense for high SKU count and dense inventory. Don't pick based on which salesperson was more persuasive.
- Standardize the format before you send the RFP. A comparison matrix beats a folder of PDFs every time.
- References need to be same-size, same-industry. "A large retailer used this" tells you almost nothing about whether it will work in your 400-person, three-site operation.
Bottom line: this project unlocked a procurement category I didn't know I needed—storage systems, not just racking. Next time, I'll have a proper spec drafted before the first sales call. This round, that lesson cost us about $650K to learn, which is the number I use in my head when I say "a storage system supplier has to earn the through-put claim in writing." At least, that's been my experience with this particular deployment.

