Operator takeaways
- Descartes buys products to run them, and is not a 3PL.
- Integrations stalled; native connectivity comes into Integration Manager first.
- The data lake is in beta, and what you build on your data stays yours.
- The runway on a mid-market WMS got longer, which changes the underwriting for anyone acquiring other operators.
- Contracts and pricing are honored as written; expect support routing changes.
The WMS is the heartbeat of any 3PL organization. I, like many of you, make years-long plans around the technology my business is built on, which is why on September 1st, when Descartes announced it had acquired Extensiv, my antennae went up. Concern isn't the right word, but I certainly wanted to learn more.
I attended the Descartes Innovation Conference skeptical and walked away feeling pretty confident that I have never had as much power as an operator to build my business the way I want to using our software partner.
I also sat down with Mikel Richardson, who runs the e-commerce division at Descartes, and David Miller, VP of Customer Growth at Extensiv, for an hour on the Supply Chain Saga podcast and asked them the questions I had been collecting from other operators since the announcement. Some of what follows comes from that conversation, and some of it comes from running a 3PL on this platform since 2020 and watching what it has done well and where it has stalled.
Listen: Supply Chain Saga Episode 22
Mikel Richardson (Descartes) and David Miller (Extensiv) with Mark Taylor. Chapter timestamps in the description.
About 1 hour
My biggest takeaways from the acquisition are as follows.
Who bought the company matters more than what they paid
The first thing I wanted to understand was what kind of buyer Descartes is, because that tells you what happens to the product over the next five years. There are acquirers who buy a software company for the customer list and let the product rot, and there are acquirers who buy a product to run it.
Descartes is the second kind, and they are built for it. They have 30,000 customers, more than 3,000 employees, and over 200 products, and they have done more than 70 acquisitions. Mikel told me they add three to five companies every year like clockwork.
They organize all of that into divisions they call pillars, and 3PL Central now sits in the e-commerce pillar alongside Peoplevox, pixi, Sellercloud, and the rest of the inventory, order management, and WMS products they own. It stays its own product line with its own brand and the same team we have been working with.

The other thing worth saying plainly, because people get confused by it, is that Descartes is not a 3PL and does not compete with any of us. They are a technology company. I bring that up because the worst version of this deal would have been a logistics operator buying our WMS and getting a window into our customers. That is not what this is.
Descartes is not a 3PL and does not compete with any of us.
I asked Mikel for an example of an acquisition that went well so I could set expectations for year one, and he pointed to OrderMine, a forecasting and demand planning company they bought earlier this year. Three things changed for those customers in the first year.
The product went from a small team handling its own R&D and support to global, around the clock support and expanded development capacity, so features ship faster. The barriers between teams selling to the same customers came down. And the customer ended up with one accountable vendor instead of a finger pointing exercise between two.

That is a reasonable template for what 3PL Central customers should expect, and it is the right order of operations.
Integrations are what stalled, and they are what gets fixed first
If you want to know what a 3PL's margin looks like, look at how long it takes to connect a new customer. When I started on 3PL Central in 2020, a Shopify integration was a scoped project. We charged the customer five or six hundred dollars and it took weeks.
Then Extensiv acquired CartRover, rebranded it Integration Manager, and almost overnight the same connection took thirty minutes if the customer had their API keys ready. That one move changed our onboarding economics more than anything else Extensiv did while I have been a customer.

Then it stopped. For roughly the last 18 months there have been no major new integration releases, and operators have been going through Techdinamics, SPS, and other partners to cover the gaps. I understand why. Extensiv was in a period where it had to batten down the hatches, and a company in that position does not spend on connectors.
So this was my first real question for Mikel, and the answer was better than I expected. Descartes already has connectivity across its portfolio that Integration Manager never built, including marketplaces, channels, EDI, and some of the older formats that still run a lot of retail. The plan is to bring that into Integration Manager rather than replace it.
Descartes operates one of the largest EDI networks in the business, and for a 3PL that currently pays a partner per connection, having that native in the WMS changes what it costs to say yes to a customer.
Bring Descartes connectivity into Integration Manager rather than replace it.
Two things I appreciated about how he answered. First, he was careful not to disrupt partners that already work. His words were that if it ain't broke, don't fix it. If you are on Techship or you use iDrive for parcel like we do, nobody is coming to rip that out.
Switching technology is the easy part now. Switching your people is the hard part, and he understands that.
Second, he said connectivity gets built where customers are asking for it, not for the sake of a longer list. That is the right discipline, and it means the fastest way to get the connector you need is to ask your CSM for it loudly.
The other half of the answer is the open API. David's point was that with AI coding tools, an operator can build an integration that would never have been profitable for Extensiv to build as a one-off.
I have seen this in practice at the conference, where a non-technical person had a working module running against the system that was built in an afternoon. It had a couple of bugs because it was an idea, not a product, but it was running, and that is the part that matters.
“Switching technology is the easy part now. Switching your people is the hard part.”
— Mark Taylor
The data is coming, and what I build on it stays mine
Anything you build on your own data is your secret sauce.
Every operator I talk to is building something right now. A pricing module, a billing tool, a report the WMS does not produce, something that pulls their numbers into a model. What has held all of that back is getting at the data without dedicating three people to it.
The data lake has been on the Extensiv roadmap for well over 18 months, and it has been one of those projects everyone wanted and nobody could quite get to. I asked where it stands.
David was candid that access to the data has existed all along through the API on a case-by-case basis, and what never happened was productizing it with pricing, packaging, and support. They are in beta with a handful of customers now and the intent is one offering that works for everybody rather than six flavors.
Mikel went further and said the direction is real-time replicas of your production database, and that Descartes coming in is going to put pressure on the business to bring that forward.
That is the piece I would watch most closely if I were you. A replica of your own production data plus an LLM is the difference between an owner who can ask a question about his business on a Tuesday night and one who has to wait for someone to build a report.
Mikel's closing advice to the audience was to get on your Claude and ChatGPT skills because data is coming, and I would take him at his word.
The question underneath that one is ownership, and I asked it directly. If I build a billing tool on my own data through their API that calculates recurring storage in a way nobody else does, is that mine, or does it end up in the product for my competitors?
The answer was about as clean as I could have hoped for. Anything you build on your own data with their API is your custom work and your secret sauce. Descartes will watch for common problems across the customer base and build generally available agents for those, and when they do, everyone gets them. They are not taking anyone's work.
For an operator who is putting real money into tooling this year, that is the assurance you need before you write the check.
“Get on your Claude and ChatGPT skills because data is coming.”
— Mikel Richardson, Descartes
On the question everyone asks about AI pricing, there is no answer yet, and I think that is the honest answer. The agent control plane they showed at the conference is early beta with no pricing or packaging.
Where AI is built into a product, Mikel wants pricing tied to things you can count, like orders and customer volume, and he said some capabilities will simply be part of the license. I would rather hear that than a number they have to walk back.
The runway got longer, and that changes the math on consolidation
This is the takeaway I did not have before the conference, and it came out of a conversation with Mikel the day before we recorded.
If you look at the WMS landscape on an invoice basis, 3PL Central sits firmly in the middle tier. Above it are the tier two systems, and above those are the tier one platforms that cost what a small warehouse costs.
The traditional path for a growing 3PL was to outgrow 3PL Central at some point and migrate up, and anyone who has been through a WMS migration knows that it is a year of your life and a real risk to your customers.
My assertion going in was that the gap between a tier one system and 3PL Central is closing, and closing fast, because the operator can now build the things that used to require the bigger system. Security and compliance are the exception, and that is where a public company parent actually helps.
Mikel's framing was sharper than mine. He said no one is better positioned than Extensiv to lengthen the time a 3PL can run on a mid-market WMS before it needs to step up, because of three things happening at once: features shipping faster, customers building their own capabilities around the system, and the Descartes network bringing tools to the table that only used to exist in tier one, like denied party screening, customs filing, and parcel and LTL connectivity through their Global Logistics Network.

So the runway gets longer, and while you still eventually grow into something heavier, because that is the nature of everything, the change moves from a change you have to make next year to one you make five years out, and that is a different capital plan.
I think about this beyond my own business too, because the one certainty in our industry is consolidation. Every time one operator acquires another, the number of WMS instances goes down by one, and I asked Mikel how Descartes thinks about that given that it shrinks their own user base.
His answer was that consolidation makes their biggest customers bigger and gives their smaller customers a way to grow and eventually sell, and that their pricing flexes whether the combined operation runs one instance or several.
I bring that up because when you are the one doing the acquiring, the WMS question is one of the first things you underwrite. An operator on a system with a long runway and an owner that welcomes consolidation is a much simpler integration than one on a system you will have to migrate off of in year one.
The unglamorous part: contracts, pricing, and security
Contracts are honored as written; data governance is unchanged.
The first thing every operator thinks when a publicly traded company buys their software is that prices are going up. I asked, and the answer was that contracts are honored as written and pricing is what you have been paying Extensiv, including the multi-node pricing where each additional warehouse is a smaller add-on to the primary license.
Mikel's reasoning was practical rather than generous: Descartes has done this 70 times and it does them no good to uproot what customers expect. If there are changes down the road they come through conversation, and he said he welcomes the feedback. That is consistent with how they have treated other acquisitions, so I take it at face value.
On security, there was work that had to happen before the deal closed, and I expected to feel it heading into peak. We did not. Mikel's point was that a public company holds a different standard than a private one, and hardening the system is part of every acquisition Descartes does.
David said the Descartes team made that process pretty painless on the Extensiv side. Keeping in mind that we are heading into the one stretch of the year where uptime is everything, that is the answer I needed.
Data governance is unchanged as well. Your data is governed by the same contract it is governed by today, and the framing Mikel used is that you own your data and they build tools that help you interact with it. For those of us who are protective of our customers' data while also wanting to do more with it, that is the right posture.
The one change you should expect to feel is operational. Over the next six to nine months, support ticket routing and some back-office processes will move as the businesses integrate.
Descartes also has single sign-on across its products today and is working toward one contract and one bill for customers who use more than one of them. Get the new support path from your CSM as soon as it exists.
What I am doing about it
Stay vocal, and just keep operating your business.
I asked Mikel my favorite question of the day, which was what a 3PL Central user should be doing over the next six months to get the most out of this, and his answer was to stay vocal.
Keep routing feedback through your CSM, give Descartes access to the power users in your building who know where the friction actually is, and tell them where they are making the right moves and where they are not.
David's addition was simpler, and I think it is the right counterweight: just keep operating your business, get your ducks in a row, and you will be ready when the new processes land.
“Just keep operating your business. We don't want you to feel anything different.”
— David Miller, Extensiv
At Warehouse Republic, that means a few specific things.
- We are asking to get into the data lake beta, because the customer-facing tooling I want to build for our brands depends on that data feed and I would rather shape the product than wait for it.
- We are building our billing automation on the API now that I know it stays ours, and I will tell Descartes what it does when it works.
- We are putting our best floor and office people in front of their team rather than filtering everything through me.
- We are going through the 200-product catalog with the specific edge cases that have cost us customers in the past, starting with EDI, to see what is already solved.
- And we are not touching anything that works, which includes our parcel setup, until the math is clearly better.
I went into September thinking I might have to revisit a years-long technology plan, and I came out of October with more confidence in it than I had before.
For operators who are thinking about what the next five years look like for their business, whether that is growing it, selling it, or figuring out which, the technology question just got easier to answer, and I am happy to compare notes with anyone working through it.
Listen: Supply Chain Saga Episode 22
The full conversation with Mikel and David is Supply Chain Saga Episode 22. It runs about an hour and the chapter timestamps are in the description, so you can jump straight to the question you care about.
