Introduction
Every Order in Mover has two locations, a pickup and a delivery, that need to be accurately geocoded before a Route can be planned and executed. When an address is received malformed, incomplete, or ambiguous, the Geocoding Providers try as much as possible to return a coordinate anyway, often with reduced precision, to avoid blocking the execution altogether. The result: drivers sent to the wrong spot, stops outside the geofence, phone chains between Driver, Contractor, and Planner, and in the worst cases failed deliveries.
Address Confidence surfaces the geocoding confidence of every pickup and delivery address directly in the Orders experience. Planners and support teams can see which Orders have a problematic address, filter for them, and take action before routing, instead of discovering the problem mid-execution.
The feature is informational: it does not block Order ingestion, planning, or execution. It gives you visibility so you can decide how to handle low-quality addresses based on your operational context.
Why Should I Use Address Confidence?
For Planners
Shift left on address problems See which Orders have low-confidence addresses before you plan them into a Route, not after a Driver calls from the wrong location. Early visibility means upstream correction becomes possible instead of live firefighting.
Bulk review instead of constant interruption Rather than reacting to address problems one by one during planning and execution, or having to review each individual Order after Route Planning, review all low-confidence Orders together, at the start of a shift, or at any cadence that fits your operation. Address Confidence is designed to support a planned review rhythm, not to flood you with alerts.
Protect route efficiency Orders with bad coordinates create late deliveries, failed stops, driver detours, and extra compensation. Catching these before a Route is built keeps planned estimates realistic.
For Customer Support
Faster root-cause diagnosis When a recipient calls about a missed delivery, you can immediately see whether the underlying address was geocoded with low confidence. That narrows down the investigation and helps you respond with accurate information.
Measurable feedback to upstream teams Address Confidence makes it visible which Orders arrived in Mover with weak addresses. This gives your organization a concrete signal to feed back to the teams responsible for data at the source (checkout, ERP, country operations).
How the Confidence Score Works
Every pickup and delivery address is scored automatically when the Order is created or updated. The score reflects how confident the geocoding provider is that the coordinates it returned actually correspond to the address you provided.
The score is derived from several signals, including:
How precisely the provider could locate the address (premise, street segment, area, etc.)
What type of entity was matched (a specific building vs. a street vs. a zip code area)
Whether the address resolved unambiguously to a single location, or multiple matches were found
What the percentage means
Address Confidence is shown as a percentage, ranging from 10% to 90%. The percentage reflects how confident we are that the returned coordinates will be close enough to the actual address, and will point the driver in the right direction.
The real-world distance associated with a given score varies country by country: a 70% score might correspond to roughly 200 meters from the true address in one country, and 500 meters in another, depending on address density, zip code size, and provider coverage.
As a rule of thumb:
90% - The provider matched the specific building, down to the street number or apartment. This is the highest confidence level and corresponds to a precise rooftop-level match.
~75% to 90% - The coordinates are still expected to be accurate at the street address level, but with a lower probability that the exact building was identified accurately.
~40% to ~74%, The match is at street level. The returned coordinates point to the street, but the exact position along it is less certain. Real-world accuracy varies significantly by country: in the Netherlands, where zip codes are small, the address is usually still within 100–200 meters; in countries like Romania, the radius can extend to 500 meters or more.
Below 40%, The provider returned the center of a zip code or an administrative area rather than the address itself. Confidence is very low, and these Orders are the most likely to cause execution issues.
The score is also affected by how many possible matches the provider found for the same address: the more candidate locations it returns, the lower the confidence.
No configuration is required, Address Confidence is active by default for all Orders.
How to Use It
Viewing confidence in the Order list
Address Confidence appears as a column in the Order list. The indicator shows the confidence of the Order's pickup and delivery addresses at a glance.
Navigate to the Orders page.
Enable the Address Confidence column from the column settings if it is not already visible.
Scan the list for low percentages.
The column is optional, you can hide it on smaller monitors or in views where it is not relevant.
Filtering for problematic addresses
The fastest way to find Orders that need attention is to filter by confidence threshold.
Open the filters panel on the Orders page.
Select Address Confidence and set the threshold (for example, show everything at or below 70%).
The list returns every Order where either the pickup or the delivery address is at or below the threshold.
This "either side" logic is deliberate: an Order with a perfect pickup but a weak delivery is still going to cause problems, and you want to see it.
Inspecting pickup and delivery separately
The Order details page shows pickup confidence and delivery confidence as separate indicators, so you know exactly which side of the Order has the issue.
Click into any Order from the list.
Under the Order's pickup section, review the pickup address confidence.
Under the delivery section, review the delivery address confidence.
Correcting the mistake
Remember to fix the mistake in your Order Management System instead of in Mover’s TMS, so that if a new Order update comes, the address will not be overwritten.
Special Cases
1. Low confidence does not mean undeliverable
A low-confidence address can still be delivered successfully, especially in countries with small zip codes or dense urban areas, where even an approximate match is within walking distance of the correct building. Address Confidence is a signal, not a verdict. Review the address on a map before deciding whether to correct it.
2. The same percentage can mean different distances in different countries
A 60% score does not correspond to a fixed distance from the true address. In a country with small, precise zip codes, a 60% match might be within 200 meters of the correct location; in a country with larger administrative areas, the same score might be 500 meters off. When calibrating your internal review thresholds, take your operational geography into account.
3. Pickup and delivery are scored independently
An Order can have a high-confidence pickup and a low-confidence delivery, or vice versa. The filter surfaces any Order with at least one weak side. Always check the Order details to see which side needs attention.
4. Corrections should generally happen upstream
If your Orders are fed into Mover from an ERP, WMS, or checkout system via integration, editing the address inside Mover will be overwritten the next time the Order is updated upstream. For customers in this setup, the most reliable correction path is to fix the address in the source system. Address Confidence helps you identify which Orders need that upstream fix.
5. The score does not change route planning behavior
Address Confidence is purely informational. Low-confidence Orders are still planned, still routed, and still dispatched like any other Order. Geofencing, optimization, and execution logic are unaffected.
6. The score can change through time for the same address
The address providers receive corrections on a daily basis, and an address that today has a confidence of 40%, can be up to 90% next time.
Best Practices
Build a review rhythm. Address Confidence works best as a planned checkpoint, for example, a quick filter pass at the start of each shift, or a weekly bulk review, rather than as a constant alert source. Create a view that automatically filters for Orders with delivery date as tomorrow, status Ready and Confidence lower than your identified threshold (70% might be a good starting point).
Calibrate your threshold to your geography. The same percentage means different real-world distances in different countries. Start with a threshold that matches your operational reality and adjust based on how often low-confidence Orders actually cause problems.
Prioritize Orders with low-confidence deliveries over low-confidence pickups. Pickups usually happen at known Distribution Centers with stable coordinates; deliveries go to unpredictable recipient addresses, which is where most problems concentrate.
Use the signal to push data quality upstream. If one of your ingestion sources consistently produces low-confidence addresses, that is feedback worth sharing with the team that owns the upstream data.
Do not treat 90% as a guarantee. A small share of failed "wrong address" stops occur on high-confidence addresses, usually because the customer entered a wrong address at checkout that the provider could still match with high quality. Address Confidence catches geocoding problems, not customer-side mistakes.
FAQ
Q: What happens if the address I provided has a very bad quality? If the address provider is not able to recognize the address at all, and does not find any match, the Order will be left in Placed status. You can filter for that status in the Order List, to ensure that all Orders are in Ready state.
Q: Can I correct the address directly in Mover? You can edit an Order's address like before. However, if the Order is updated from an upstream system after your edit, your correction will be overwritten. For stable fixes, correct the address at the source.
Q: Can I manually pin a location on the map or override the GPS coordinates? Not yet, manual coordinate override is not part of this release. Reach out if you would find it useful.
Q: Does Address Confidence apply to Routes or Services? Address Confidence is currently exposed on Orders only, where we expect it to have the most impact.
Q: Why is the maximum score 90% and not 100%? Even under ideal conditions, there is always some residual uncertainty in geocoding, the provider cannot fully guarantee that a matched coordinate corresponds to the exact location the customer intended. Capping the score at 90% reflects that inherent limit and prevents a false sense of certainty.
Q: Will Mover block low-confidence Orders from being planned? No. Address Confidence is informational by design. Many low-confidence addresses are still fully deliverable, and blocking would create operational friction without a net benefit.
Q: How often is the score updated? The score is calculated when the Order is created or when its address changes. It reflects the geocoding result at that moment.
Q: My company sends GPS coordinates to Mover together with the Order. What does the percentage mean in this case? The score measures the quality of the address, not the quality of the GPS coordinates you provide. Your Route will still be planned and executed using the coordinates you sent, a low confidence score does not degrade that. What the score tells you is how reliably the address text on the Order resolves to a real location, which matters whenever someone (a driver using their own navigation, a support agent looking up the delivery) works from the address rather than from your GPS.
Key Takeaways
Address Confidence surfaces the geocoding confidence of every pickup and delivery, directly in the Orders list and Order details, as a percentage between 10% and 90%.
The percentage reflects match quality from the address provider, not a fixed distance, the same score can correspond to different real-world accuracy depending on the country.
A filter lets you find Orders with weak addresses, on either the pickup or delivery side, before they reach planning or execution.
The feature is informational and non-blocking: it gives Planners and Customer Support visibility, without changing how Orders are routed.
Best results come from a planned review rhythm, a threshold calibrated to your geography, and pushing persistent data-quality issues back to upstream systems.