Descartes Alternatives for Forwarders, Brokers and Trade Teams: Keep the Network, Replace the Operations Layer

Why logistics teams start looking

The first reason is that the estate stopped feeling like one product. Descartes has grown substantially by acquisition, which is public and not a criticism in itself, since acquisition is how a company assembles customs content, carrier connectivity, routing, and forwarder systems faster than anyone could build them. The practical consequence for a customer is that different modules were built by different teams with different assumptions, so the seams between them are places where your data gets rekeyed or reconciled.

The second reason is transactional pricing meeting a growing business. When you pay per document, per filing, per user, or per shipment, growth reprices you automatically. That is defensible for a service that carries real regulatory cost. It stings when the volume growth is in low margin freight or when a large customer's document count is out of proportion to what they pay you.

The third is that your operating model is more specific than a product can be. Forwarders differentiate on how they quote, how they handle exceptions, what a customer sees, and how quickly a file turns into an invoice. Those are the parts a general system covers generically, and the parts where a spreadsheet or an email chain usually appears alongside the software.

What Descartes genuinely does well

The connectivity and content are the product, and they are hard to replicate. Filing to customs authorities across jurisdictions, keeping up with schema and rule changes, maintaining denied party and sanctions lists, and holding tariff classification content current is a permanent maintenance job with a compliance consequence when it slips. Paying a vendor to carry that is a rational allocation of risk.

The carrier and partner network is the second genuine asset. Connections to airlines, ocean carriers, brokers, and trading partners have network effects: the value comes from everyone else already being connected. A new entrant cannot conjure that, and neither can you.

Breadth is the third. For a company that needs customs, screening, routing, and visibility, one vendor relationship is simpler than five, with one procurement cycle and one accountable party when something breaks. If you are mid sized, running standard freight flows, and your compliance team sleeps well, staying is the right answer and this page should not talk you out of it.

Where it actually strains

The first strain is the join between modules. Where an acquired portfolio has been integrated at the data level, it works well. Where it has not, operations staff bridge the gap, and bridging looks like exporting a file, reformatting it, and importing it somewhere else. That work is invisible in a demo and enormous in a year.

The second is configuration ceilings in operations. Quoting rules that reflect your commercial reality, job costing that recognises revenue and cost accruals the way your finance team wants, exception handling that matches how your desks actually work: these are the areas where products offer configuration and businesses need specificity. You adapt, and the adaptation shows up as procedures your staff carry in their heads.

The third is reporting across the estate. The questions that matter are cross module by nature: margin per file by trade lane, exception cost by customer, which accounts consume the most operational hours against what they pay, duty spend against drawback recovered. Answering those means data from several places joined somewhere neutral, which is a warehouse project you own whichever vendor you use.

Your realistic options

Option one is to consolidate rather than replace. Many trade teams run a Descartes module beside a competitor's tool and an internal spreadsheet because of how procurement happened over ten years. Rationalising to fewer moving parts, with a proper data flow between them, is cheaper than any migration.

Option two is another platform. CargoWise from WiseTech is the dominant comparison for forwarders wanting one deeply integrated operational system, at the cost of a demanding implementation and a strong opinion about how you should work. Magaya suits smaller forwarders and warehouses. On the trade compliance side, Thomson Reuters ONESOURCE Global Trade, E2open, and MIC are the usual alternatives for classification, screening, and filing. For visibility, project44 and FourKites specialise where suite modules generalise. Switching to CargoWise in particular is a real strategic choice: you gain integration and lose flexibility, and both are genuine.

Option three is the hybrid: keep filing and screening connectivity from a specialist, and build the operations layer you actually compete on. This is the pattern that fits most independent forwarders, brokers, and importers with distinctive processes.

Whichever route you pick, get your operational data into a warehouse first. It is cheap relative to any migration, it answers the reporting questions that usually triggered the search, and it turns the decision into evidence rather than opinion. Teams that do this often discover their real problem is two accounts consuming a wildly disproportionate share of operational hours, which is a commercial conversation rather than a software purchase.

When a custom build pays back

Build when your niche is your business. Perishables, project cargo, pharmaceutical cold chain, oversized freight, and bonded warehousing all have workflows a general system approximates. If your reputation rests on handling something unusual reliably, the software that runs it should reflect the way you handle it.

Build when the customer portal is a sales asset. Shippers increasingly choose partners on visibility and document access. A portal designed around your customers' questions, with their documents, their milestones, and their exception history, is a retention tool. Rented portals look like everyone else's.

Build when duty and cost recovery deserve their own analytics. Drawback, foreign trade zone inventory, first sale valuation, and reconciliation programs turn on data you already generate. The recovery opportunity is often large enough to fund the tooling that finds it, but only if the data from filings and operations is joined and queryable.

Two counter signals are worth stating plainly. If you file a few hundred entries a year, your operational cost is people rather than software, and better process beats better tooling every time. And if nobody in your business will own the process design, a custom system will encode the current confusion at higher resolution and charge you for the privilege.

What you should not rebuild

Do not rebuild customs filing connectivity, denied party and sanctions screening, or tariff classification content. These change constantly, carry penalties when wrong, and are available as services. Do not rebuild carrier EDI where a network already offers it, since the value of connectivity is the other side being there. And do not build your own accounting: a forwarder platform should push costs and revenue into a real ledger, not become one.

Migration reality

The rule is that open files do not move. Let shipments in flight complete in the old system and start new files in the new one, because a partially documented entry is exactly the wrong thing to migrate. That means running both for a cycle, which for ocean freight can mean a couple of months.

Historical records carry retention obligations. Customs records must remain retrievable for years after entry, so keep an archive that a compliance officer can search under audit pressure, not a folder of exports nobody has opened. Test filing connectivity in a controlled way before you depend on it: a rejected entry is an operational event with a customer waiting on the other end. And plan retraining around your desks, since operations staff have process knowledge that lives in habit, and a new screen sequence slows a good clerk down before it speeds them up.

Cost bands and the honest recommendation

Vendor pricing here is typically per user plus per transaction or per document, with content subscriptions on top and implementation services in year one. Your negotiating room at renewal is usually volume commitment rather than list price. On the build side, from Digital Heroes delivery experience: a custom operations and customer portal layer sitting over existing filing and screening services runs $60k to $150k over 12 to 18 weeks. A full forwarder or trade operations platform with quoting, job costing, document automation, portal, and integrations runs $180k to $400k. Budget 15 to 20 percent of build cost a year for changes, because trade rules and customer demands both move.

The honest recommendation: keep the regulatory connectivity, always. Consolidate before you migrate. Switch to a deeply integrated suite if standardising your operation is genuinely what you want. Build the operations and customer layer when your niche, your service, or your duty recovery is what earns your margin, because that is the part no vendor will ever model as well as you can.