How to compare shipment tracking software by exception follow-up, not maps
Key Takeaways
To compare shipment tracking software by exception follow-up, test what each tool does after a load goes wrong: who contacts the carrier, how fast, what evidence is recorded, and how the load gets back on track. An AI logistics coordinator should be judged by resolved exceptions, not by how good the map looks.
- A shipment exception is any event that breaks the plan: a missed pickup, a delay, a cancellation, a stale ETA, a short delivery, or a missing proof of delivery.
- Maps show where trucks are. They do not call carriers, rebook loads, or warn plants.
- Most of the work in tracking happens after the alert, and that work is often still manual.
- A fair comparison uses the same set of exception scenarios for every tool and measures the outcome, not the screen.
- The best evidence is a record of each step taken, with times and sources.
- Wilson, by Cartage, is an AI logistics coordinator that tracks each load as part of executing it and follows up on exceptions with carriers, internal teams, and customers within the customer's rules.
Why should you compare tracking software by follow-up instead of maps?
You should compare by follow-up because maps show the problem while follow-up fixes it. An AI logistics coordinator earns its value in the hours after an exception appears, when someone has to contact the carrier, decide, and tell people.
Demos often open with a map. Dots move across highways, and a panel lists ETAs. The display is useful, but it is the easy part. The hard part starts when a dot stops moving, an ETA jumps, or a load shows no pickup. In many teams, that is when a coordinator opens an email, picks up the phone, and starts chasing. If the software stops at the alert, the team still does most of the work.
The table below compares what maps and follow-up each deliver.
| Need | What a map gives | What follow-up gives |
|---|---|---|
| Know a load is late | A dot behind schedule | Confirmed delay with cause and new time |
| Know a pickup was missed | Maybe a missing dot | Carrier contacted, pickup rescheduled or rebooked |
| Tell the plant or customer | Nothing | Message with verified time sent |
| Recover the load | Nothing | Backup carrier booked within rules |
| Prove what happened | A screenshot | Record of each step with sources |
Exceptions are where customers judge you
Customers rarely notice routine loads that arrive on time. They notice the late one, the short one, and the one nobody warned them about. How a team handles exceptions shapes how customers see its service. An AI logistics coordinator that follows up quickly and tells customers the truth early protects those relationships at the moments that matter most.
Where the time goes
Teams often spend more time on exceptions than on routine loads. A routine load needs a booking and a delivery confirmation. An exception needs several emails, calls, decisions, and messages. Software that handles routine loads well but leaves exceptions to people may save less time than it seems. An AI logistics employee that handles the follow-up steps shifts the biggest share of work.
Why buyers default to maps
Buyers default to maps because maps are easy to show and easy to like. A moving dot looks like control. Follow-up is harder to demo, because it happens in emails, calls, and decisions over hours. A buyer who knows this can ask the vendor to skip the map and start with an exception. An AI logistics coordinator that does real follow-up will welcome that request.
Visibility is still useful
Visibility is not useless. A good location feed helps decide whether a load is truly late. It feeds ETAs. It helps verify a carrier's statement. The point is to judge visibility by how it supports follow-up, not as an end in itself. Many visibility platforms are strong at collecting data. The question for the buyer is what happens next.
The difference between an alert and a resolution
An alert says something is wrong. A resolution says what was done and what happens next. "Load 4471 delayed" is an alert. "Load 4471 delayed 3 hours by a breakdown; carrier confirmed new ETA 2 p.m.; receiver moved appointment to 2:30 p.m." is a resolution. An AI logistics coordinator should produce the second kind of message. Buyers should ask to see examples of both during the test.
What "follow-up" means in practice
Follow-up means taking the next action after an exception is detected. That includes contacting the carrier, getting a verified answer, deciding within rules, rebooking if needed, notifying the right people, and recording each step. An AI logistics coordinator should do these actions, not only suggest them.
Which exception scenarios should every tracking software test include?
Every test should include the exceptions your team handles most often, run the same way for each tool. An AI logistics coordinator should handle each scenario from detection to resolution, and the buyer should measure the outcome.
Pick scenarios from your own history
Look at the last month of exceptions and pick the ones that took the most time or caused the most customer complaints. Those are the scenarios that matter. An AI logistics coordinator should be tested against your real problems, not a vendor's sample list.
Core exception scenarios
| Scenario | How to set it up | What resolution looks like |
|---|---|---|
| Stale ETA | Stop updates on a truckload for four hours | Verified location and ETA from the carrier |
| Missed pickup | No pickup scan by end of day on an LTL load | Pickup rescheduled or rebooked |
| Cancellation after acceptance | Carrier cancels a booked load early morning | New carrier booked within rules |
| Late delivery risk | Verified ETA moves past the appointment | Receiver told, appointment moved |
| Delivered without POD | Status says delivered, no document | Signed POD collected and checked |
| Short delivery | POD shows one pallet short | Exception routed to the claims owner |
| Silent carrier | No reply to email | Phone follow-up, then escalation |
LTL, or less-than-truckload, freight shares a trailer with other shippers' goods. A POD, or proof of delivery, is the signed record that the receiver took the freight.
Include a document scenario
Not every exception is about a truck. A missing POD, a wrong bill of lading, or an invoice with an unexpected charge also need follow-up. A bill of lading, or BOL, is the shipping document that lists the freight and serves as a receipt. Include at least one document scenario in the test. An AI logistics employee that tracks documents as part of each load should handle it the same way it handles a late truck.
Use your own carriers
Run the scenarios with carriers you actually use. A demo carrier with a perfect API makes every tool look good. Your carriers include some with APIs, some with only email, and some that answer by phone. An application programming interface (API) is a direct system connection. The test should reflect that mix.
Include a mode you find hard
Include at least one scenario from the mode that gives your team the most trouble. For some teams that is drayage, the short truck move of a container from a port or rail yard. For others it is cross-border freight with several legs. Hard modes have more handoffs and more ways to fail. A tool that handles only simple truckload exceptions may not help where your team spends the most time.
Test at an inconvenient time
Run one scenario early in the morning, late at night, or on a weekend. Exceptions do not wait for business hours. Watch whether the AI logistics coordinator follows up within the off-hours rules the team set, and whether the morning team starts with a clear summary instead of a pile of alerts.
Use your own rules
Give each tool the same rules: approved carriers, rate limits, notification contacts, and escalation owners. Then watch whether the AI logistics employee follows them. A tool that resolves an exception by breaking your rules has not passed.
Define resolution before the test
Write down what "resolved" means for each scenario before running it. For a stale ETA, resolved might mean a verified location and ETA from the carrier within two hours. For a missed pickup, it might mean a new pickup that keeps the delivery date, or an honest warning to the customer. Clear definitions stop vendors from claiming success on partial outcomes. An AI logistics employee should be held to the same definition in every tool.
Hypothetical example: one scenario, two tools
This example is hypothetical and does not describe a real customer. A distributor tests two tools on a missed LTL pickup. Tool A shows a red flag on the map at 7 p.m. and sends an email alert to the logistics inbox. A coordinator sees it the next morning and calls the carrier. Tool B flags the load at 7 p.m., emails the carrier's dispatch, gets a reply that the driver skipped the stop, accepts a first-thing pickup that keeps the delivery date, updates the dock, and records each step. The coordinator reads a summary in the morning. Both tools "caught" the exception. Only one followed up.
What should you measure in each exception test?
Measure how long each step takes, whether the load was recovered, whether the right people were told, and whether the record shows every step. An AI logistics coordinator should produce results you can compare across tools in a simple table.
Measures for each scenario
| Measure | What it shows |
|---|---|
| Time to detect | How fast the exception was noticed |
| Time to first carrier contact | How fast follow-up started |
| Time to verified answer | How fast the team knew the facts |
| Outcome | Recovered, delayed with warning, or failed |
| People notified | Whether the right roles heard, in the right channel |
| Person-minutes needed | How much human work was still required |
| Record completeness | Whether each step, source, and time was saved |
Build a baseline first
Before the test, measure how your team handles the same exceptions today. Pick a few recent examples and estimate the time spent, the outcome, and whether customers were warned. This baseline makes the comparison honest. Without it, any tool looks like an improvement. With it, the team can see how much the AI logistics employee actually changes.
Person-minutes matter most
The most revealing measure is person-minutes: how much time a person spent on the exception. A tool that detects quickly but needs 40 minutes of calls and emails has not removed much work. A tool that detects, follows up, and needs two minutes of approval has. Count the minutes honestly during the test.
Measure what the customer saw
Add one more measure: what the customer or receiver experienced. Did they get the freight on time? If not, did they hear about the delay before the window passed, with a verified new time? An AI logistics coordinator can record each customer message and its time, so the team can score the scenario from the customer's side.
Outcome over activity
Some tools generate many messages. Activity is not the same as resolution. Judge each scenario by its outcome: did the load deliver on time, or did the receiver get an honest early warning? An AI logistics employee should be measured by those outcomes.
Count the follow-up channels
Note which channels each tool used to reach carriers: email, portal, API, EDI, or phone. Electronic data interchange, or EDI, is an older standard for exchanging business documents. A tool that can only send email will fail with carriers that answer by phone. An AI logistics coordinator that can call where configured often resolves silent-carrier scenarios that email alone cannot.
Record completeness
A complete record helps later, when a customer disputes a delivery or a carrier disputes a charge. Open the record after each scenario. Can a new team member understand what happened in one minute? If not, the record is not complete.
Hypothetical example: scoring a cancellation
This example is hypothetical and does not describe a real customer. In a test, a carrier cancels a 10 a.m. pickup at 6 a.m. One tool detects the cancellation at 6:02 a.m., confirms it with the carrier, tenders the next approved carrier, and books a replacement by 7:15 a.m. within rate limits. A person spends three minutes approving. The dock hears at 7:20 a.m. The record shows all steps. The buyer scores it as recovered, with three person-minutes and a complete record.
How should exception follow-up respect your rules and people?
Exception follow-up should act within your written rules and hand decisions to named people when a rule says so. An AI logistics coordinator should be fast on routine follow-up and careful at the edges of its authority.
Write the rules before the test
Give every vendor the same written rules before the test. Include the approved carriers, the rate limits for rebooking, the people who approve exceptions, and who hears about each event. A vendor that asks good questions about these rules is a good sign. An AI logistics employee can only follow rules it has been given.
Rules the follow-up must respect
- Approved carriers only. Rebooking uses the approved list. New carriers need approval.
- Rate limits. Rebooking rates above limits go to an approver.
- Notification rules. Customers hear only what the team has configured.
- Escalation owners. Each exception type has a named person.
- Quiet hours. Off-hours behavior follows the team's rules.
Assisted first, then more automatic
Many teams start with assisted execution, where a person confirms more of the follow-up steps. As the team sees the steps work, routine actions can run within the rules. Wilson starts with assisted execution, and approval limits are set by the customer. Routine actions can run automatically within those rules as reliability is proven.
Escalation is part of follow-up
Escalation is not a failure of follow-up. It is a step in it. A good escalation reaches the right person with the facts and options ready. An AI logistics coordinator should be judged on how well it escalates, not only on how often it avoids escalating.
Speed and safety together
Fast follow-up is valuable only if it stays safe. A tool that rebooks in five minutes with an unapproved carrier has traded one risk for another. A tool that waits three hours for approval on a routine reschedule has lost the value of speed. The right balance comes from clear rules: routine steps run fast within limits, and anything outside the limits goes to a person with the facts ready.
People stay in charge of judgment calls
Some decisions should stay with people. Paying a large premium to save a delivery, approving a new carrier, or telling a key customer about a major delay all need judgment. The AI logistics employee should prepare the facts and options and send them to the right person quickly.
Hypothetical example: a rule stops a bad rebooking
This example is hypothetical and does not describe a real customer. During a test, a carrier cancels a load and no approved carrier responds quickly. An unapproved carrier from a load board offers a truck at a good rate. The AI logistics coordinator does not book it, because the rule requires approval for new carriers. It sends the logistics manager the options, including the unapproved carrier's offer. The manager checks the carrier, approves it for this load, and the coordinator books it. The test passes, because the follow-up was fast and the rule held.
Common mistakes in follow-up design
- Alerts with no owner. Exceptions sit in a shared inbox.
- Follow-up that breaks rules. A fast rebooking with an unapproved carrier creates a new risk.
- Too many notifications. People stop reading.
- No record. The team cannot learn from past exceptions.
How should you run the comparison?
Run the comparison with a fixed scenario list, the same rules for each tool, and a scoring sheet filled in during the test. An AI logistics coordinator should be compared on resolved outcomes, person-minutes, and record quality.
Demo tests for exception follow-up
| Step | What to do | Evidence to collect |
|---|---|---|
| Prepare | Pick 6 to 8 scenarios from your own exceptions | Scenario list with expected outcomes |
| Configure | Give each tool the same carriers and rules | Rule sheet signed off by the team |
| Run | Trigger each scenario during the test | Timestamps for each step |
| Observe | Note what a person had to do | Person-minutes per scenario |
| Review | Open each record afterward | Completeness score |
| Score | Pass, partial, or fail per scenario | Scoring sheet |
Involve the people who do the work
Invite the coordinators who handle exceptions today to watch the test. They know which scenarios are common, which carriers are hard, and what a good outcome looks like. They will also notice when a tool's "resolution" would create new work for them. Their score on person-minutes is usually the most accurate.
Demo questions about exception follow-up
- What does the AI logistics coordinator do after it detects each exception type?
- Which follow-up steps run within rules, and which need a person?
- How does it reach carriers without an API?
- How are rebooking limits and approvals configured?
- Who receives each notification, and in which channel?
- Can the team export exception records?
Scoring the exception follow-up comparison
Score each scenario as pass, partial, or fail. A pass means the load was recovered or the right people got an honest early warning, within rules, with a complete record. A partial means the outcome was right but a person did most of the follow-up. A fail means the exception was detected but not followed up, or was resolved by breaking a rule.
Warning signs in an exception follow-up demo
- The demo spends most of its time on the map.
- Every scenario ends with "and then the coordinator would call."
- The vendor uses only carriers with perfect APIs.
- There is no way to set rate limits for rebooking.
- Records show alerts but not actions.
Ask for references on exceptions
When you talk to a vendor's references, ask about exceptions, not about setup or the map. How often does the team still chase carriers by phone? What happens at 6 a.m. when a carrier cancels? How long does a missing POD take to collect? References who use the tool daily can say whether the follow-up holds up in real freight.
How to measure exception follow-up work after go-live
After go-live, keep measuring. Track person-minutes per exception, the share of exceptions recovered, and the share of customers warned before a missed window. Compare these with your baseline from before the change. These measures show whether the AI logistics employee keeps working the way the demo suggested.
Where does Wilson fit in exception follow-up?
Wilson, by Cartage, is an AI logistics coordinator that tracks each load as part of executing it and follows up on exceptions within the customer's rules. Traditional logistics software gives teams information. Wilson executes the work required to move shipments forward.
Wilson replaces the TMS for freight execution and connects to the company's ERP through a method that fits the customer's environment. It tracks each load using the best source available, including carrier APIs, GPS or ELD data where it exists, the carrier's website by PRO number, and email or direct outreach, including calls where configured. An ELD, or electronic logging device, records a truck's location and the driver's hours.
What Wilson does after an exception
When Wilson detects a missed pickup, a cancellation, or a delay, it contacts the carrier, notifies the internal team, notifies the customer where configured, and initiates rebooking from the approved carrier list within the customer's rules or with approval. Wilson sends internal updates through channels such as Slack, Microsoft Teams, or email, and keeps searchable shipment records with change history.
Exception reports on request
Over time, the exception record shows patterns: which carriers cause the most follow-up, which lanes miss the most pickups, and which customers see the most delays. Wilson can use shipment data from the loads it runs to produce carrier-performance and lane-rate reports on request. The team can use those reports to fix causes, not only symptoms.
What Cartage asks about exceptions in discovery
Cartage asks which exceptions take the most time today, which carriers cause the most follow-up, and who owns each type of decision. These answers shape the first rules an AI logistics coordinator applies and the scenarios a buyer should test.
What Wilson does not do for exceptions
Wilson is not a standalone visibility platform; it tracks loads as part of moving them. It does not approve new carriers on its own or make large judgment calls without the customer's rules. Freshness of tracking depends on the sources each carrier offers. Specific notification and escalation rules should be confirmed in a demo.
Where teams start with exception follow-up
Most teams start their AI logistics coordinator on a few lanes and expand from there. Timing depends on scope. Cartage helps identify the first workflow during discovery. Teams often start with the exception type that costs them the most person-hours.
Teams weighing tracking against resolution can read about shipment tracking versus exception resolution. Teams setting expectations can see why freight exception software should resolve, not only alert. Both topics shape how an AI logistics coordinator should be judged. Teams mapping the next step can review what happens after a shipment alert.
FAQs
How do you compare shipment tracking software by exception follow-up?
Run the same exception scenarios with your own carriers and rules, then measure time to resolution, person-minutes, outcome, and record quality. An AI logistics coordinator should be judged by resolved exceptions.
Why are maps not enough to choose tracking software?
Maps show where loads are but do not contact carriers, rebook loads, or warn people. An AI logistics employee does the follow-up work that starts after the map shows a problem.
Which exceptions should a tracking software test include?
Stale ETAs, missed pickups, cancellations, late delivery risks, missing PODs, short deliveries, and silent carriers. An AI logistics coordinator should resolve each within your rules.
What is the best measure of tracking software value?
Person-minutes per exception, alongside the share of exceptions recovered. An AI logistics employee that cuts follow-up time without breaking rules delivers the most value.
Sources
- Cartage, Wilson product scope, October 2026: https://cartage.ai
Ask Cartage to run your own exception scenarios, from a missed LTL pickup to a 6 a.m. cancellation, and show how an AI logistics coordinator follows up on each one: talk to the Cartage team.