Logistics Software M&A: How AI quietly changes the Deal Now

Descartes Systems Group bought two logistics software companies in eight days. Tai, a transportation management platform for freight brokers, cost about US $100 million on 24 August. Extensiv, a warehouse management system for third-party logistics providers, cost about US $120 million on 1 September. Descartes paid for both from cash on hand. Both are logistics software M&A deals of an ordinary kind, but the reason for them is not.

Logistics software M&A is not the story, because it has been running for as long as I have been active in the supply chain business. What changed is the reason for it. Acquirers are increasingly buying transaction data as well as applications, since software that takes over a decision needs a long record of people making it. So the buying continues. The European manufacturer signing a three-year TMS contract this quarter is contracting for a product and a counterparty. Either may have changed by renewal.

What logistics software M&A buys today

Neither firm needed artificial intelligence to start doing logistics software M&A. Descartes gave its reason in its own announcements. The Tai release describes the value added as transaction, carrier and shipment execution data flowing into the Descartes Global Logistics Network. The Extensiv release is blunter. Extensiv reaches across sales channels, marketplaces and carriers, so it produces omnichannel fulfilment data. The release says plainly that this data exists to feed AI capability. Descartes bought a warehouse management system. The release then spends its space on what that system generates. That is the order, inventory and fulfilment record of several thousand logistics operations.

That is a different logic from the one driving logistics software M&A five years ago. Buying an application bought a customer base and a revenue line. Buying a data set buys an input to a model, and that input can outlast the application that produced it. So logistics software M&A changes what a target is worth. What matters is not the installed base alone but what it generates. Descartes describes Tai as orchestrating the whole shipment lifecycle: quoting, carrier sourcing, load execution and billing in one workflow. A million shipment records are useful. A million records linking the quote, the carrier chosen, the execution and the outcome are worth far more. That is exactly the kind of training set a system needs if it is later going to make those choices itself.

The same move seen from the product side

WiseTech makes the point from the other end. Its FY26 results, released on 26 August, put revenue up 79% to US $1,395.9 million, carried mostly by the e2open acquisition. CargoWise revenue grew 11%. But one figure matters more to a buyer. More than 95% of CargoWise customers have moved onto CargoWise Value Packs. Most of that 11% came from the commercial model change rather than from freight volume.

Then the framing. WiseTech now calls its platform a system of record and execution. Its chief executive told The Loadstar that CargoWise is becoming an AI “system of execution”. Described that way, the vendor stops selling a record of what happened and starts selling the decision itself. WiseTech is building a model for e2open that ties price to usage.

The two firms move toward the same model from different directions. Descartes widens the data and the workflows inside its network. WiseTech widens the share of execution its platform performs inside those workflows. The more decisions the software makes, the more the execution data underneath is worth. It also becomes easier to price against activity rather than against users.

A regulator has already tested this

The consequence of this consolidation for buyers is not a surprise. Australia’s competition regulator ran the test in December 2025 and published the result the following month. The ACCC accepted a court-enforceable undertaking requiring WiseTech to divest Expedient. Expedient arrived with e2open and had competed with CargoWise there. The regulator’s stated concern was direct. WiseTech already holds substantial market power in logistics software. The acquisition removed a constraint the two products placed on each other. Without the divestment, the ACCC found, WiseTech could raise prices or cut service quality.

Two things follow for a European customer. First, the reasoning is relevant outside Australia, because the same vendors operate globally. It does not transpose automatically, since market definition, competitors and customer alternatives differ by jurisdiction. Second, the remedy does not apply for Europe. The undertaking restores Expedient as an independent competitor in Australia and New Zealand alone, and no European authority has run the equivalent review.

Two things that complicate it

The first is that the capability being bought has not yet shown it works. BCG surveyed 30 large logistics companies this year. Ninety-seven per cent called AI a strategic priority. Thirteen per cent reported measurable financial impact. Eighty per cent named a single standalone tool as their most useful AI. So the acquirers are not buying data because the execution layer already pays. One way to read this is as a possibility for AI economics that have not yet arrived. The acquirer takes the workflows and the data before their value in automated execution is settled.

The second is the market. WiseTech counts 24 of the 25 largest global freight forwarders and 47 of the 50 largest third-party logistics providers among its users. That is the logistics provider market. It is also where both of this fortnight’s deals sit, since Tai is a broker system and Extensiv a 3PL warehouse system. A manufacturer buying its own transport or warehouse software is in a different market, with independent vendors still in it.

So the exposure splits. Where the manufacturer buys directly, bargaining power survives and the terms below are winnable. Where it hires a forwarder or a 3PL, the lock-in is real and it belongs to somebody else. A manufacturer hires a forwarder, the forwarder runs its business on one of these platforms, and the platform’s pricing arrives inside the forwarder’s rate.

What a TMS contract should carry now

Four terms do the work, and none is about the price.

Assignment and change of control comes first. Most TMS contracts permit assignment to a successor entity without consent. That clause is what hides the change of counterparty until the invoice arrives. Consent, or at least a termination right on change of control, costs nothing to ask for before signature. After signature the buyer’s power to get it falls sharply.

Data rights on exit come second. They are not a copy of the records, which every vendor grants. The real question is whether the vendor may retain, aggregate and commercialise the transaction history after termination. The practical test is one question. Can the vendor use a customer’s shipment history to improve a model whose output it then sells to another customer, including a competitor? A data clause that does not answer that is incomplete. So a TMS contract silent on derived data leaves one of the relationship’s most valuable assets undefined.

Third is the pricing basis rather than the price. A move from seats to transactions changes the unit. A discount negotiated on the old unit can then become largely irrelevant. Fix the basis for the term, or price the conversion in advance while there is still something to trade.

Fourth is continuity of the product. An execution layer takes over workflows the buyer’s own staff run today. Service levels then have to attach to the quality of decisions and to a right of review, not only to system availability.

What to do about it this quarter

Assume logistics software M&A will have moved the market again before renewal. Look at the renewal date on your TMS contract and your WMS contract, then at the assignment clause and the data provisions. Renewal inside the next eighteen months means the negotiation starts now, while the incumbent still wants it. Then put the same four questions to your forwarder and your 3PL. Their software choices are the half of your exposure you neither control nor see.

If a proposal is on your desk this quarter, budget for the bill after the pricing basis changes, not for the license quoted today. On the evidence of the past fortnight, the application is only part of what logistics software buyers are paying for. The increasingly valuable asset is the data that runs through it.

MVAventures reviews supply chain technology and compliance business cases for manufacturers, wholesalers and logistics operators. If a proposal is on your desk and you want to know which line it actually moves, contact us .