System Integration in Freight Forwarding

A mid-sized Brazilian freight forwarder touches six to eight different systems on a single import shipment. The ERP where the file is opened. The carrier portal for booking confirmation and container tracking. Siscomex Web for the CE Mercante (Brazil’s electronic cargo manifest). The Portal Unico for the DUIMP (Brazil’s single import declaration). The terminal website for cargo arrival status and pickup scheduling. The municipal tax system for issuing the NFS-e (electronic service invoice). Email for receiving pre-alerts and documents from overseas agents. WhatsApp for sorting out exceptions with customs brokers and clients. Each system has its own login, its own data structure, and no automatic connection to the others. The operator is the one who ties them together, copying and pasting fields from one screen to another. When system integration depends on the operator, operations scale by headcount, and margin shrinks accordingly.

The operator as the integration layer

At most Brazilian freight forwarding companies, the operator sits in front of a dozen browser tabs. They check the booking on the carrier portal, copy the container number, paste it into the shipment file in the ERP, go back to the portal for the updated ETA, open Siscomex Web to transmit the CE Mercante, return to the ERP to record cargo data, open the inbox to download the commercial invoice the overseas agent sent, cross-check the fields against the system, and open one more tab to fill in the service invoice on the municipal tax portal.

Each step takes seconds. Across 30, 50, or 80 shipments per month, those seconds become hours. In our experience working with freight forwarders in Brazil, operators spend between 15 and 30 minutes per shipment on tasks that amount to copying a field that already exists in one system and pasting it into another. At a forwarder handling 400 shipments per month with six operators, that comes to 100 to 200 hours per month of work that disappears once the systems talk to each other.

Accuracy suffers too. Every time the operator types a field that already exists in another system, there is a chance of error. A container number with one wrong digit in the CE Mercante triggers a correction filing. A gross weight mismatch between the bill of lading and the DUIMP triggers an inspection channel. An incorrect taxpayer ID on the NFS-e triggers a rejection and a returned invoice. These errors show up as rework, penalties, and delayed collections. No budget line says “lack of integration,” but that is where the money goes.

We have written about parallel spreadsheets as a symptom of a system that does not deliver what the operator needs, and about key-person dependency as a risk when operator knowledge is not recorded. Missing system integration feeds both problems: the operator maintains the spreadsheet because they need to pull data from different sources, and they accumulate knowledge in their head because only they know the sequence of screens that makes a given shipment work.

Five connections that determine whether your system works or just records

Not every integration carries the same weight in a freight forwarder’s daily routine. Five connections change the way operators work when they are in place. Without them, the operator bridges the gap by hand.

Start with Brazil’s Portal Unico and Siscomex. The CE Mercante, the CCT (the new Cargo and Transit Control system replacing Siscarga), the DUIMP, and the DU-E (export declaration) all require transmission of data that already exists inside the shipment file. When the forwarder’s system transmits the CE Mercante directly, with the house bill of lading, consignee, and cargo data pulled from the file, the operator only reviews and confirms. Without this connection, the operator opens Siscomex Web, types every field, cross-checks the ERP screen alongside, and hopes no digit was transposed along the way.

Next, carrier and overseas agent portals. Bookings, container tracking, ETAs, and routing updates come from carriers. Today, many freight forwarders check these data points manually on each shipping line’s portal. When the system pulls the information via EDI or API, the operator stops switching between portals and the ETA shows up inside the shipment file without anyone going to look for it.

Then terminals and bonded warehouses. Cargo arrival status, discharge dates, storage deadlines, and pickup availability come from terminals. At ports like Santos, Paranagua, and Itajai, the lookup is still manual at many forwarders. The operator accesses the terminal website, searches by container number or bill of lading, notes the result, and updates the file. With integration, those dates enter the system as the terminal publishes them, without intervention.

Municipal electronic invoicing (NFS-e) is another high-value connection. Every freight forwarder in Brazil issues electronic service invoices, and each municipality has its own system. The operator or billing clerk who opens the city tax portal, fills in the invoice fields from the shipment data, and then records the NFS-e number back in the ERP is performing a manual integration. When the system issues the NFS-e from the shipment’s billing data, the clerk reviews the amount and authorizes the invoice without leaving the screen.

Finally, inbound document processing. Pre-alerts, commercial invoices, packing lists, carrier invoices, and quotes arrive by email, and each document contains fields the operator will retype into the system. Retyping data from incoming documents is one of the most repetitive tasks in the operation. With automated extraction, the document’s fields enter the shipment file with a confidence score and the source page, and the operator reviews only what the system flagged as uncertain instead of typing everything from scratch.

How to measure the cost of an isolated system

A freight forwarding owner who wants to size this cost can start with three numbers the operation already has.

Count the duplicate entries per shipment. Take a typical ocean import FCL file, from booking to settlement, and count how many times the same field appears across different systems. The container number, for example, gets typed in the ERP, in the CE Mercante, in the terminal lookup, in the DUIMP (if not integrated), in the NFS-e, and sometimes in the free-time tracking spreadsheet. If the same field is typed across four or five screens, each entry is a chance of error and a fraction of a minute that, multiplied by hundreds of shipments, turns into days of work.

Measure time spent outside the system. Ask an experienced operator to log, over one week, how much time they spend on screens other than the ERP: carrier portals, Siscomex Web, terminal websites, the NFS-e portal, email with documents, WhatsApp. The number tends to surprise. In our experience, operators at mid-sized freight forwarders spend between 30% and 50% of their day in systems that do not talk to the ERP. So the system the owner invested in captures, at best, half of the operator’s workday.

Calculate the cost per retyping error. Tally how many CE Mercante corrections, how many invoices returned for incorrect data, and how many customs classification discrepancies occurred in the past three months. Most of these errors originate in the manual transfer from one system to another. A CE Mercante correction has a deadline and a penalty risk set by Brazil’s Federal Revenue Service. A returned invoice delays collection by one to two weeks. A discrepancy in the DUIMP can change the customs inspection channel and delay cargo release.

Put it together: a forwarder handling 400 shipments per month, with 20 minutes of avoidable retyping per shipment and an operator cost of BRL 35.00 per hour, spends BRL 4,666 per month on redundant data entry alone. If 2% of those shipments produce a retyping error that costs an average of BRL 500 in corrections, rework, and delayed collections, that adds another BRL 4,000. The total exceeds BRL 100,000 per year, not counting the opportunity cost of an operator who could be resolving exceptions instead of copying fields.

Integration that works starts with the process, not the API

The usual reaction when a freight forwarding owner realizes the operation is fragmented is to ask the system vendor for more integration. The conversation turns technical fast: API, EDI, webhook, web service, XML file. These words describe how integration is built. They say little about what needs to be integrated first.

Integration priority should follow the volume of data the operator moves by hand, not the technical complexity. If the forwarder’s operation is mostly ocean import FCL through Santos and Itajai, the integration that saves the most is the one that eliminates retyping at the highest-volume touchpoints: the CE Mercante, terminal lookups, and inbound documents. If the forwarder does a lot of exports, the DU-E and shipping document generation carry more weight.

Start with the operator’s workflow. Follow a shipment from start to finish and note every time the operator leaves the system to look up or enter data somewhere else. Sort the list of exits by frequency. That sorted list is your integration roadmap. The top three items on it probably account for 70% to 80% of the time lost to retyping.

A common trap is chasing integrations that look impressive in a report, like real-time container tracking on a map, when the operator is still typing the bill of lading number twice per shipment. The highest-return integration removes repetitive work from the operator who sits at the desk, on every shipment, every day. The shipments per operator metric improves when integration eliminates dead time between systems, because the same operator can handle more shipments in the same day.

How Tier2 Cargo connects freight forwarding operations

The integrations described in this article are part of the Tier2 Cargo workflow. The CE Mercante and the CCT are transmitted from the shipment data, without the operator opening Siscomex Web. The NFS-e is issued from the billing step, on each branch’s municipal system. Documents arriving by email go through AI-powered field extraction: the system reads the pre-alert, commercial invoice, or carrier invoice and suggests the fields for the operator to review instead of retype.

The operator works on one screen, inside the shipment file, and sees data from other systems arrive as the operation moves forward. Shipments per operator go up because the time spent copying data between screens goes away.

Learn more about Tier2 Cargo or get in touch.

To size the problem in your own operation, count how many times the operator types the same field across a single import shipment and how many different screens they work in. Those two numbers show how much of your operation still depends on someone copying and pasting, and where integration would pay off first.