Blog

10 signs that you should fire your Image Recognition vendor

Image Recognition has the potential to reduce the average visit time at approximately 30%, freeing up the field reps to get more visits done and to improve the in-store execution. At the same time, the rep can capture the entire category during a regular visit, helping category managers to tap into the full potential of the store data.
This is a solid technology that keeps on improving based advances of AI and mobile device computing power, attracting more and more entrepreneurs to provide solutions in this space.
Since Image Recognition acts at the core of the field force, influencing both the bottom and top line, it is paramount the selected vendor is able to deploy a successful project.
Causes of project failure can be a combination of technology, services and change management. Below we describe 10 points to be aware of when procuring an Image Recognition vendor.
  1. Proven track record: it is important to select vendors with solid examples in your vertical.
  2. Technology: many vendors claim to use technology but end up doing manual recognition by leveraging cheap labour from emerging markets;
  3. Accuracy: should be above 90% (human accuracy is around 70%);
  4. Speed: visits results should come back to the field reps in near real time so they can improve the store execution;
  5. Scalability: the vendor should be able to recognize thousands of SKUs for thousands of reps visiting thousands of stores concomitantly.
  6. Price: there are vendors that charge costs per taken picture, making it difficult to expand visit coverage and store penetration. Ideally, the price model should be either per visit or per user;
  7. KPIs: standard KPIs and possibility to create custom scores;
  8. Reporting: standard reports and customizable views;
  9. Integration: public API available to integrate with legacy systems;
  10. Standalone app: a robust standalone app that includes modules for visiting stores, checking in, receiving calculated KPIs and reports.
Besides these points, it is important the vendor has a clear project methodology, governance model, support structure and SLAs. With all these points covered, you will be set for a successful project.

About Inspector Cloud
Market-proven and award-winning Image Recognition based on AI. Our solution empowers elite CPG brands of the caliber of Mars, Nestlé and Unilever across several categories and thousands of SKUs.
More information at https://inspector-cloud.com/