Somewhere between 200 and 300 products call themselves a warehouse management system. Almost all of them will answer yes to almost every question on a standard requirements list, and none of them will be lying.
Jeff Hamilton, a partner at LIDD, had me on It's the End of the Week! in April 2026 for a Playbook episode on selection. He implements these systems for a living, which made it a more useful argument than the ones I usually have about this.
Every quote block below is me, on the episode, condensed for reading.
Critical Levers
The term carries a lot of weight in how the Sprint gets described, and it deserves a plainer definition than it usually gets:
When I'm walking their floors, no matter how big or small the business is, there's usually a small core set of what I call critical levers. These are workflows, or features, that if they don't exist in the system, this client can't use it. They're going to fall down. These are core to their business.
A critical lever is a requirement that disqualifies. Its absence ends the conversation with that vendor, whatever the rest of the scorecard says.
Most operations have a handful, and the handful is what selection should be built around. The example I keep reaching for comes out of food and beverage:
Lot genealogy. You get a lot of vendors out there who say yeah, we support this, but it's surface level. There's no teeth behind it. One person sees it and thinks one thing, the other thinks something else. The vendors aren't evil, they're incentivized to get the signature, so they're not going to dig in. Maybe the system tracks a date and some kind of text field that's the lot number. Nothing behind it.
Ten Questions Instead of a Thousand
I'm very opinionated about this. I don't like RFPs, because they prize the wrong thing. They're looking at this gigantic laundry list of feature comparisons. I've authored them, I've read them, I've answered them, and by and large it's just false confidence.
Jeff made the sharper version of the point, from the answering side of the table. Ten complex questions tailored to one operation are far harder to dodge than a thousand generic ones. A short answer looks bad when there are only ten of them, and a vendor staring at a genuine gap in question six has to weigh whether they want the delivery problem that comes with saying yes.
A long requirements list is easy to answer honestly and impossible to answer revealingly. Cutting it to ten removes the vendor's cover without accusing anyone of anything.
The rest of the feature list still matters, and it's largely settled. Table stakes get most operations 80% of the way there, which is exactly why running an RFP as the opening move spends the most effort on the least decisive part.
The Implementation Ratio
The most practical thing in this conversation is a number I hadn't put on the site anywhere:
The way this gets framed is against your year one subscription fee. An implementation ratio of 1.5x to 2x has been fairly standard for a long time. Get to 3x the subscription and now you're into the heavy ones. It's a way to gauge and grade what's going on.
Both ends of that range tell you something. 3x and up usually means a Swiss Army knife system that does everything and needs configuring into your shape. Well under 1.5x is the more dangerous reading, and Jeff has watched it land:
A low number looks like efficiency to a finance team or a CEO who has never done one of these. To anyone who has, it reads as a vendor with very little implementation structure, which means the hours didn't disappear. They moved onto your team, and the question is whether your team knows that and is technical enough to absorb it.
An implementation quote states how much of the work the vendor intends to do. A low one moves the work rather than removing it.
The Vendor Would Rather Sell Software
I'll just stake a claim. WMS software vendors are not in business to implement the software. They do it because it's complicated. Ultimately they'd rather be in their high margin software business.
That makes a real partner ecosystem worth reading as a signal rather than a footnote. It suggests the vendor is mature enough that other firms have invested in learning the product, and it gives you somewhere to go when the fit with one implementer is wrong. Fit is mostly about disposition:
There are folks that are very technical, so they come in with the lens of, we're going to mold your operation around the software. And then you have the flip side where folks are more operational: we'll look for a system that fits your operation without having to reinvent it. There's a balance there, because it's never pure play one or the other.
Two implementers with identical product knowledge will hand you different operations at the end, because one of them is solving for the software and the other for the building.
Jeff's addition was the one I'd want a buyer to hear, since he sells these engagements. The person who can translate between the vendor's object names and your team's actual vocabulary is worth more on a project than the person who can fix the device that stopped scanning. Sales cycles are built to show you the second one.
The Plan Took Them a Year
I had a client last year, about Super Bowl time. I went out, ran through this sprint, gave him reports. His quote to me was that the quick wins were worth the investment all by themselves. The things I called out gave him permission to act on what he already suspected was important, but that had become normal.
Then the part that matters more:
It took them almost another year to actually work through what I provided. Small team already wearing a bunch of hats. Some of the things that needed doing took time. And they were also trying to grow and transform from one type of business into a proper 3PL.
Writing the plan is the easy half. The operations that most need one have the least room to execute it, and that gap is where readiness work actually lives.
That's the honest shape of this work, and it's why the deliverable is built to be executed by the client rather than only by me. They got there, selected a system, and moved. It took a year, and pretending otherwise would make the plan look better and the buyer worse prepared.
