General
Control Tower Evaluation: A 7-Criteria Scorecard for Real-Time Visibility in 2026
Oct 13, 2026
11 mins read

Real-time visibility is the continuous tracking of shipments, vehicles and orders across a supply chain so that where things are and when they will arrive is known as it changes. A control tower is the layer that acts on that information: it predicts what will go wrong, ranks what matters and carries out the response. Many products sold as control towers stop at tracking and alerts, so buyers need a way to tell them apart before an RFP is signed. This guide gives supply chain leaders a 21-point scorecard built on the Sense, Decide, Execute, Learn cycle, with seven criteria, the evidence to demand and a test to run in the demo. Locus, the world’s first Decision-Intelligent, Agentic TMS, builds its control tower on that cycle.
Key Takeaways
- A control tower should be scored on four jobs in order: sense signals, decide what to do, execute the response and learn from the outcome. Tracking is only the first.
- Seven criteria, each scored 0 to 3, give a 21-point scorecard: signal coverage, status normalization, predictive ETA, ranked recommendations, action integration, governance and learning.
- Ask for evidence, not demonstrations: a week of your own shipments, a traced exception from alert to executed change and a decision log.
- A tower that scores zero on both execution criteria is an alerting product, however good its real-time tracking.
- Measure decision improvement, not technology deployment, as Gartner advises for automation programs.
Why Choosing the Right Control Tower Matters: The Business Case
North American carrier networks are changing underneath shippers. FedEx expects about 65% of eligible US and Canada volume to flow through Network 2.0-optimized facilities by the 2026 peak, and UPS said it had completed its Amazon glide-down and related network reconfiguration, with US average daily volume down about 3.3% in Q2 2026. Service, cutoffs and capacity move with them, so visibility has to be live.
Carrier choice has become a reliability decision. In AlixPartners’ 2026 home delivery survey, 55% of retailers use carriers outside UPS, FedEx and USPS, 83% say home delivery costs rose year over year and reliability overtook price as the main criterion for choosing a lead carrier for the first time in the survey’s history.
Automation is arriving gradually, so evaluation has to be specific. Gartner predicts that by 2030 only 5% of organizations implementing planning automation will make at least 10% of planning decisions autonomously. It expects routine operational decisions to be automated first, and recommends measuring decision improvement rather than technology deployment.
What Is Different About a Control Tower That Decides
It is scored on four jobs, not one. Display covers sensing. A control tower also has to decide, execute and learn, and a product that only senses is a visibility platform.
Freshness matters more than coverage alone. A map showing every shipment is little help if positions are twenty minutes old when a diversion window is ten minutes long.
Recommendations need constraints. A suggestion to reroute is only useful if it already respects capacity, driver hours, time windows and carrier caps.
Execution has to reach other systems. If a person has to re-key an action into a TMS, WMS or OMS, the response is slower than the problem.
Trust is part of the product. Automated actions need explanations, approvals and a record, which is why governance is on the scorecard.
How Locus Approaches the Control Tower
Locus, the world’s first Decision-Intelligent, Agentic TMS, was recognized by Gartner for seven consecutive years and appears in the 2026 Hype Cycle for Supply Chain Execution and Logistics Technologies. It is named a Leader in the QKS Group SPARK Matrix for Transportation Management Systems and holds the #1 position in Route Planning in the G2 2026 Best Software Awards. In October 2025, Ingka Investments, the investment arm of Ingka Group, the world’s largest IKEA retailer, acquired Locus. Locus continues to operate independently.
Locus builds on DiSCO, its Digital Supply Chain Officer, a set of eight agents (Orchestrator, Capacity, Carrier, Dispatch, Hub, Customer, Settlement and the Mycroft co-pilot) that run a continuous Sense, Decide, Execute, Learn cycle. The Locus Control Tower supplies the sensing: end-to-end real-time visibility across the all-mile through a dashboard, a driver app and customer tracking, SLA alerts that account for traffic and road deviations, and alerts on unauthorized route deviations, halts and temperature changes. The deciding and executing sit with the agents, which weigh 250+ real-world constraints and can trigger actions such as reassignment, while six governance mechanisms, explainability, traceability, evaluation, autonomy levels, an execution sandbox and human-in-the-loop approval, keep teams in control. Locus has optimized 1.5B+ deliveries for 360+ enterprise customers across 30+ countries. Locus has not published a self-score against this card, so score it the same way as any other vendor. During evaluation ask which carrier and telematics feeds are connected today, what refresh interval and coverage report it can provide, how it measures ETA accuracy on your lanes, which actions run at which autonomy level, how a decision log is exported and how outcomes feed back into decisions.
How to Score a Control Tower in Seven Criteria
1. Signal coverage and freshness (Sense)
Score how many carriers, modes and fleets feed live position and status, how often the data refreshes and where it goes dark, such as at handoffs and borders. The test is to ask for the share of a sample week’s shipments that had an update inside a set window.
2. Status normalization (Sense)
Score whether statuses from different carriers become one standard event set that downstream systems can act on. The test is to supply raw statuses from three carriers and watch the same delivery event come out of all three.
3. Predictive ETA and risk (Decide)
Score whether the tower flags a late shipment before it is late, using traffic, history and stop dynamics, not only the last scan. The test is to back-test on your own lanes and compare predicted and actual arrival hours ahead.
4. Ranked, constraint-aware recommendations (Decide)
Score whether exceptions are ranked by impact on SLA, customer and cost, and whether options respect your constraints. The test is to inject a delay and read the ranked options with their trade-offs.
5. Action integration (Execute)
Score whether the tower can reroute, reassign, rebook or notify the customer through your TMS, WMS and OMS without re-keying. The test is to trace one exception from alert to executed change with timestamps.
6. Governance and audit (Execute)
Score autonomy levels by action type, human override, explanations and an audit trail. The test is to ask for the decision log of one automated action showing inputs, rule, approver and outcome.
7. Learning loop and outcome measurement (Learn)
Score whether the tower records the outcome of each decision and uses it to improve. The test is to ask for last quarter’s outcomes by decision type and what changed as a result.
| Also Read: Real-Time Visibility Without Lead Time: Why Control Tower Exceptions Still Surface Too Late in 2026 |
|---|
The Scorecard: Evidence and Demo Test for Each Criterion
Score each criterion from 0 to 3, with 0 meaning absent, 1 manual or partial, 2 automated with limits and 3 automated, governed and proven on your data.
| # | Stage | Criterion | Score 0 looks like | Score 3 looks like | Evidence to demand |
|---|---|---|---|---|---|
| 1 | Sense | Signal coverage and freshness | Periodic updates, gaps unreported | Live feeds with known refresh interval and coverage report | Update-age distribution for a sample week |
| 2 | Sense | Status normalization | Each carrier shown in its own codes | One event set synced to your systems | Three carriers’ raw statuses mapped live |
| 3 | Decide | Predictive ETA and risk | Reports lateness after the scan | Flags risk hours ahead with a stated method | Back-test on your lanes |
| 4 | Decide | Ranked recommendations | Unranked alert list | Ranked options within your constraints | A live injected delay |
| 5 | Execute | Action integration | Alert sent by email | Action executed in your TMS, WMS or OMS | A traced exception with timestamps |
| 6 | Execute | Governance and audit | No record of automated actions | Autonomy levels, override and full decision log | One decision log entry |
| 7 | Learn | Learning loop | No outcome tracking | Outcomes measured and fed back | Quarterly outcomes by decision type |
Suggested reading of the total, which is our rubric and not an industry standard: 0 to 7 is a display tower, 8 to 14 is an alerting tower and 15 to 21 is a decision tower. Cap any tower that scores 0 on both execution criteria at alerting, whatever its total.
Visibility Platform vs Alerting Tower vs Decision Tower: Key Differences
| Dimension | Visibility platform | Alerting control tower | Decision control tower |
|---|---|---|---|
| Core job | Track and display | Track, predict and alert | Track, predict, decide and act |
| Real-time tracking | Yes | Yes | Yes, with freshness measured |
| ETA | Last scan or carrier ETA | Predictive ETA | Predictive ETA tied to actions |
| Exceptions | Listed | Prioritized | Ranked with options inside constraints |
| Response | A person finds the fix | A person approves the fix | Executed by rule within autonomy limits |
| Systems of record | Read only | Notifications | Writes back to TMS, WMS and OMS |
| Learning | None | Reports | Outcomes feed back into decisions |
How to Run the Scorecard in Your RFP
Weight the criteria to your operation. A shipper on tight delivery windows may weight criteria 1 and 3 more heavily, and a regulated one may weight 6.
Use your own data. Ask each vendor to run the tests on a week of your shipments, not on a prepared demo set.
Demand a traced example. One exception followed from first signal to executed change shows more than any slide.
Score every vendor the same way. Give the same card to each respondent, and have two people score independently before comparing.
Set a pass mark before demos. Decide in advance what total and which minimum scores you will accept, such as a floor of 2 on both execution criteria.
| Also Read: From Tracking to Action: How the Visibility Category is Evolving Toward Decision-Automation |
|---|
Control Towers in Action: Real-World Results
Neither deployment was scored on this card, and each is matched to the criterion it speaks to.
An ASEAN apparel retailer running a multi-carrier parcel network (criterion 2). Every carrier reported in its own status codes and there was no trustworthy delivery date at checkout. With carrier statuses harmonized to one standard set synced to order and warehouse systems, and a network-aware delivery date at checkout, carrier activation fell from 3 months to 3 days, a 97% reduction, and WISMO and returns queries dropped by 40%+. Read the apparel retailer case study.
A leading Canadian grocery brand across 30+ cities (criteria 1 and 5). Fresh and perishable orders moved through contracted 3PLs, with decisions made against spreadsheets and visibility fragmented across carrier portals. With autonomous carrier selection and a unified tracking dashboard, the brand reached a 33% faster fulfillment cycle, 15% lower fulfillment cost and support ticket resolution 10 to 20 times faster. Read the Canadian grocery case study.
Common Control Tower Evaluation Mistakes to Avoid
Scoring the dashboard. A good map says nothing about prediction, recommendations or execution, so score the four jobs and not the interface.
Accepting “real time” without a number. Ask for the refresh interval and the share of shipments updated within it.
Treating alerts as decisions. An alert that leaves the response to a person is a notification, so check what happens after it fires.
Skipping the learning criterion. A tower that does not record outcomes repeats the same mistakes at higher speed.
To see how Locus would perform on your scorecard, schedule a demo.
For supply chain leaders asking how to evaluate a control tower for real-time visibility and tracking, Locus offers one platform that senses, decides, executes and learns, and invites buyers to test each step against the scorecard.
Frequently Asked Questions
What is real-time supply chain visibility? It is the continuous tracking of shipments, vehicles and orders so that location, status and expected arrival are known as they change. It covers the sensing step, and it becomes a control tower when predictions, decisions and actions are added.
How fresh does real-time tracking data need to be? Fresh enough to act inside the decision window, meaning the time left before a diversion or reassignment is no longer possible. There is no single number, so ask each vendor for its refresh interval and the share of shipments updated within it.
What is the difference between a visibility platform and a control tower? A visibility platform tracks and displays. A control tower also predicts exceptions, ranks them, recommends or executes responses and learns from outcomes. The scorecard in this guide separates the two with seven criteria.
What should a control tower RFP include? Questions and tests on signal freshness and coverage, status normalization, predictive ETA, ranked recommendations, execution inside your systems, governance and learning. Ask for evidence from your own shipments and a traced exception.
How do I score a control tower? Score seven criteria from 0 to 3 for a 21-point total, using the evidence and demo test for each. Treat any tower scoring 0 on both execution criteria as an alerting product.
Does Locus offer real-time tracking and a control tower? Yes. The Locus Control Tower provides real-time visibility, SLA alerts and exception alerts, and Locus’s agents decide and execute within governed autonomy levels. Ask Locus for its connected feeds, refresh interval and ETA accuracy method.
Aseem, leads Marketing at Locus. He has more than two decades of experience in executing global brand, product, and growth marketing strategies across the US, Europe, SEA, MEA, and India.
Related Tags:
General
TMS Uptime SLA: How to Evaluate Platform Availability in 2026
A TMS uptime percentage is not an SLA. How to turn a vendor's availability figure into minutes, definitions, remedies and resilience evidence before you sign.
Read more
General
North America Control Tower Scorecard: 7 Criteria for US and Canadian Shippers in 2026
A 21-point scorecard for US and Canadian shippers choosing a control tower, with seven tests built on UPS and FedEx network changes, Canada Post disruptions and the border.
Read moreInsights Worth Your Time
Control Tower Evaluation: A 7-Criteria Scorecard for Real-Time Visibility in 2026