Automated Photo Compliance Software

How OR’s AI features are improving compliance and execution accuracy.

Many companies are using artificial intelligence (AI) to aid in store compliance. Companies like One Door’s Merchandising Cloud software uses “AI to filter out the dreaded associate selfies and poor photos.”

With the launch of APC clients are seeing some clear benefits both at HQ and at the store, such as:

  • An increase in planogram execution accuracy of 80%
  • A 10% reduction in Level of Effort (LOE) to confirm and communication planogram execution.
  • 50% reduction in HQ Level of Effort in not having to manually review images.
OR Illustration Mobile Print Order Colourful

At Optimum Retailing (OR), we go a step further. We use AI to filter photos as they are taken to confirm if a fixture matches the planogram found in OR. We call this our Automated Photo Compliance feature, or APC. Stores no longer need to manually confirm and take a picture to upload for manual approval. Instead, OR uses deep learning to train our system to recognize if a store has placed the correct products and signage on the correct fixture. If not, the system will reject the store’s execution and advise them of what is missing. A simple guided tour through the use of augmented reality (AG) shows each store what products they need to place on the fixture, and the AI engine confirms that the execution matches the planogram from HQ. It’s that simple.

The result is an increase in planogram execution accuracy in-store, and a reduction in the level of effort by HQ staff from not having to manually review thousands of images to confirm that stores have executed the correct planogram or request new photos to be taken if the planogram is incorrect.

The feedback OR has received from early tests of APC indicate that stores love the guided tours conducted through AG and the simplicity of both the tool and process, while HQ loves the time and money they save.

Read How OR is helping bricks and mortor retailers up their game

Related Insights: