AI Visibility: Solving Phantom Inventory & Display Scoring
16/11/2026

I heard AI Visibility can score displays, but how is it different from manual counting?
Just last week, I was sitting in a meeting room at a large FMCG company in Ho Chi Minh City. Their modern trade channel manager complained: "We have inventory data from our sales system, but when we check the store, the shelf is empty. How can we sell what isn't there?" This is the phantom inventory problem that many businesses are facing.
People often think manual counting is sufficient, but reality shows that about one-third of checks are inaccurate due to human fatigue or bias. AI Visibility changes this. It turns display scoring into an automated, objective process. Instead of an employee taking photos and guessing, the system analyzes the image immediately to see if the product is in the right position and in sufficient quantity.
So, what specific information does a shelf photo provide?
This is the part people are most interested in. When a photo is fed into VTraks, the system doesn't just see "a shelf section." It identifies each specific SKU.
- Determines the position of each product on the shelf.
- Scores planogram compliance: whether this product meets or fails the standard position.
- Checks for out-of-stock items on the shelf (out-of-shelf detection).
- Records competitor products and prices next to yours.
This process happens extremely fast. According to vtraks.com, recognition accuracy reaches 99.97% and processing time is under 1 second per image. This means data from thousands of stores can be aggregated to a central dashboard in real time, instead of waiting for end-of-week Excel reports.
What happens if a sales rep sends a screenshot or fake photo to cheat?
This is the question I get most often from sales managers. In the past, detecting fake photos was nearly impossible. You could re-photograph an old image, edit it with a photo editing app, or simply take a screenshot from another phone.
In practice, when applying specialized AI Vision for image quality control, the system can detect anomalies such as duplicate images, screenshots, or unnatural light distortions. VTraks integrates this capability to ensure that the data you see on the dashboard is authentic. There is no room for "pretty reports" while the shelf is empty.
Do we need to change our sales team's workflow?
Very little. That is the strength of modern AI solutions. The sales team still works as usual: visit stores, take photos of shelves with their phones, and send them in.
The difference lies on the backend. Previously, they had to sit and count each product and enter the data. Now, AI does that work. Employees receive immediate feedback: "This shelf is missing 3 bottles of orange juice in the corner position." They can restock immediately or report the issue. The process becomes simpler, but accuracy is many times higher.
What is the difference between AI Visibility and AI cameras in factories?
People often confuse them because both use the term "AI Vision." But the application context is completely different.
AIV Camera (aivcamera.com) focuses on security and internal operations. It turns existing IP cameras into AI cameras for face recognition, PPE monitoring, or counting people entering and leaving. The goal is safety and access control. According to aivcamera.com, this system has 99% accuracy and alerts in under 1 second, currently operating on over 1,000 cameras.
In contrast, VTraks focuses on out-of-home market data. It doesn't care who you are, but rather whether your product is in the right place, at the right price, and how it competes on the customer's shelf. One monitors people and assets inside, the other monitors products and the market outside.
If we want to deploy AI Visibility, what infrastructure do we need to prepare?
It is best to prepare very little. This is a major advantage of SaaS platforms in this field. You don't need to buy expensive servers or install complex software on employee computers.
VTraks runs on the web. Sales reps use smartphones to take and send photos. Data is processed on the cloud. You just need to access the dashboard to view reports. For businesses with high data security requirements, or those operating in markets with strict data regulations like Vietnam, Mexico, or Thailand, choosing a provider with stable infrastructure and clear security commitments is a key factor. AIVISION, as a leading AI solution developer in Vietnam, always prioritizes this element in every technical design.
How can I know if my business is affected by phantom inventory?
There is a simple way to self-check. Take the average inventory report from your sales system (ERP/POS) for the past month. Then, randomly select 10 stores and ask your sales team to send actual photos of those key products.
Compare the inventory numbers on paper with the actual quantity on the shelf. If the discrepancy is too large, and you cannot explain the reason (e.g., theft, expired goods recalled but not entered), then you are likely losing revenue due to phantom inventory. At that point, AI display scoring is not just an option, but a tool to save profit. Ask yourself: Do you trust your inventory numbers? If not, it's time to look at the shelf more clearly.
There is more here than one article can hold. Keep reading on the AIVISION blog, look at our display scoring solution, or get in touch.