Flow assumptions · interface ownership · acceptance evidence
Simulation LabDefine a Site Assessment
← Alstef Group engineering blog

How to Evaluate WMS Integration Manufacturers: Three Scenarios, Three Different Playbooks

2026-11-20 · Jane Smith · Warehouse control integration

The "Just Compare Quotes" Trap

If your first step in evaluating WMS integration manufacturers is a spreadsheet with line-item pricing from every vendor, one of those vendors will get you eventually—the only question is which one, and in what way.

This is something I deal with every year, usually right when a project is late and someone senior asks why the integration was harder than expected. My job is to review deliverables, integration proposals, supplier tender responses, and compliance documentation before they go out. On paper the numbers look clean. The differences live in what "integration" actually covers, and who is on the hook when the handoff fails.

It's tempting to think you can rank integrations by price per site and call it a day. But the evaluation criteria that matter shift depending on project shape—not brand, not group size. Apply the wrong playbook to the wrong scenario and you won't necessarily pick the worst vendor, you'll pick the least-fitting one. Those are two very different failures.

Here are three scenarios I've seen enough times to have a working approach for each. They aren't exhaustive, and most operations sit somewhere between them. But if you can identify which one you're closest to, the right questions get a lot clearer.

Scenario A: First Pilot, or a Single Site With Modest Volume

This is the "I hope the big vendor will take us seriously" scenario. It's also where the normal rulebook gets thrown out most often.

In my opinion, modest volume does not mean you should accept diluted integration. A 500-square-meter warehouse with a botched WCS-to-WMS handoff is just as expensive to fix as a 50,000-square-meter one—smaller blast radius, same failure mode. I've seen a pilot that took a vendor's junior team two weeks to "stabilize" and then ran into exception handling gaps for months afterward.

Three things to actually look at:

First, who owns change management. Not configuration—training, for when the pick team's daily motion changes. Second, how the WCS/WES layer handles unplanned exceptions, because pilots throw off exceptions constantly and they expose how much design thought actually went into the edge cases. Third, never accept "that's standard practice" as an answer. A pilot needs a WES supplier willing to discuss boundary conditions, not one eager to present a concept deck.

The surprise usually isn't the software license. It's the rework time when someone discovers in week five that the original handoff design doesn't survive contact with the actual exception rate. Pilots have a lower tolerance for failure than larger rollouts, not a higher one—there's no backup plan yet. I'd argue that's the whole point of running one.

And yes, small orders shouldn't be treated as one-off transactions. The pilot client often turns into next year's multi-site client asking for a bulk warehouse execution system—which no one can predict with confidence. But if a supplier is treating you by short-term volume, you'll feel it in the first two months, and you'll remember it at the next tender.

Scenario B: Replacing an Existing WMS/WES at Mid-Size

The first misconception here is that "the WMS is always the hard part." That was true about a decade ago when most mid-size facilities were manual or semi-automated and the main concern was inventory accuracy. Today, in many mid-size operations, the genuinely difficult layer is the edge—the point where WMS instructions hand off to WCS and machine motion. I'd argue that's where integration evaluations are most often misweighted. Teams spend 80% of their time comparing WMS feature grids and 20% on integration, and then the integration is what blows up.

What to evaluate instead: migration path, parallel-run strategy, and data quality from the legacy system. Migration tooling has to work with what you actually have, not the idealized format from a demo environment.

The gut-vs-data moment I remember most clearly: a spreadsheet said go with one integrator whose quote was toward the low end—unit economics looked fine. Something about the integration scope and go-live support terms felt off. Turned out their migration tooling assumed a data structure materially different from the legacy system's, so most of the cut-over was manual cleanup. The savings didn't survive the first four months. No hard data proving this is a pattern, but I'd test the tooling against your messy data before signing.

Also, don't over-standardize in this scenario. If the existing site has quirks—non-aligned dock doors, mixed-batch aisles, that one refrigeration zone nobody wants to talk about—an experienced integrator will bake them into the migration plan. If the demo is all textbook flows, that's a flag. It may not surface until after go-live.

Scenario C: Multi-Site or Bulk-Throughput Deployments

Real scalability isn't a feature count. It's what happens when you're running three sites at once and you need per-site version control, a single upgrade cadence, and documented deviations from the standard.

This is where OEM and private-label flexibility becomes a very practical variable. A multi-site deployment doesn't always need to run the same system at every site—but it does need a clear boundary between what's standardized and what's customized, and a contract that keeps that boundary from becoming a custom build at every site.

This is also where a lot of WES suppliers for bulk warehouse execution systems fail—not at the concept level, but at per-site execution discipline. For companies like Alstef Group—a French-headquartered automation group operating across WCS/WES/WMS integration and material handling—the multi-site evaluation often comes down to support model, not feature set. Does the Alstef Group headquarters in the Bois-d'Arcy area sit inside the upgrade cycle, or is it outsourced per site? Who certifies a local site build? Those questions don't get answered in the demo, and they're the ones that will hurt most.

The other thing I watch: whether the supplier will commit to per-site acceptance testing, not a paper walkthrough. Multi-site rollouts can't afford a full go-live at every location. A supplier who insists "validation is only possible through full go-live" has quietly told you their deviation process isn't ready.

How to Tell Which Scenario You're In

Three questions, before you get distracted by feature grids:

1. If you swapped WMS tomorrow, could your warehouse run for a day?
2. How many exception types do you handle per shift, on average?
3. Will throughput be meaningfully higher 18 months from now?

If the answers are "yes, almost none, no"—you're in A. Prioritize integration documentation and a WES supplier willing to negotiate boundary conditions. Don't over-engineer; you're not buying a multi-site platform yet.

If they're "no, lots, no"—you're closer to B. Reframe the evaluation around migration, data quality, and parallel-run support. Watch for over-standardization in the demo.

If they're "no, lots, yes"—you're in C. Add one more dimension: whether the supplier's support model actually reaches each site, or whether it's a regional rep pretending to be one.

If you're hovering between two scenarios, I'd evaluate against the more complex one. The cost of over-preparing is usually lower than the cost of discovering the integration decision was scoped for the wrong scenario.

And one last thing: don't drop evaluation standards because the project is small. The worst integration problems I've reviewed weren't on flagship deployments—they were on the early 12-pallet pilot where the vendor sent whoever was available and it took two years to unwind.


Leave a technical comment

Do not include operational credentials, personal data or confidential facility information.