How To Do Real-Time Image Recognition With ShelfWatch

image recognition

Large FMCG companies conduct regular shelf analysis to validate their in-store execution. Hundreds of thousands of images are collected monthly across different parts of the country. It is difficult to analyze such a large number of images manually. And in order to reduce the burden of manual labor, various large brands have automated the process with the help of image recognition technologies.

Challenges in Manual Analysis

Manual analysis is complicated, time-consuming, and costly. It is difficult to manage around work-peaks. Manual analysis is often prone to human error, insufficient analysis & theoretical interpretations. All this makes it difficult to ensure the quality of retail audit data.

Keeping all these redundancies in mind, Karna AI has leveraged the power of image recognition and object detection technologies to create ShelfWatch. ShelfWatch is an automated retail shelf monitoring tool that helps retailers achieve the desired in-store execution and compliance.

Real-time image recognition with ShelfWatch

image recognition

Karna AI’s data collection app, StreetSnap, plays a significant role in achieving real-time image recognition with ShelfWatch. StreetSnap is an integration app for data collection to analyze images for retail solutions. The app is able to give near real-time retail KPI feedbacks using image analysis and deep learning. StreetSnap makes it easy to collect images that are directly transferred to the ShelfWatch cloud/algorithm that detects POSM and SKUs.

Street snap makes it easy to collect images that are directly transferred to the ShelfWatch cloud/algorithm that detects POSM and SKUs. It analyzes SKU along with Geotag, shop name, and date, giving a complete analysis of in-store execution.

StreetSnap helps FMCG brands by solving various bottleneck issues such as:

Human Bias: Street snap gives objective, standardized, and complete analysis.

Reliability: Control the data quality during the collection of data with complete transparency.

Speed & Cost: Low cost and research turnaround time make it scalable.

Advantages of StreetSnap

  • Save the time of your sales representatives
  • Reduction of staff and expenses for its maintenance
  • Increase in sales
  • Instant response to the situation inside the retail outlet
  • Effective use of POSM
  • Timely placements of the products on the shelf
  • Reduces share of empty shelves or out of stock

How StreetSnap uses image recognition to give real-time insights

image recognition

The working of StreetSnap is a simple three-step process.

Step 1: Sales reps stand in front of the shelf and capture images of the shelf, and upload the images easily.

Step 2: The images are sent to the ShelfWatch cloud for analysis.

Step 3: Within minutes, sales reps get actionable reports on their mobile device and the management team receives a detailed analysis of the performance report.

StreetSnap’s salient features

  • The UI is extremely user-friendly. Multiple users can do data collection for the same project, making the process of data collection quick and uncomplicated.
  • Softwares such as StreetSnap require an active internet connection for data upload. However, with StreetSnap, images can be clicked even in a no-internet zone without hindrance and can be uploaded once an internet connection is available.
  • StreetSnap solves the major problem of data quality because it works on real-time image quality check. In the case of blurry images, StreetSnap notifies the sales rep for the same.
  • Smart features like angle or eye-level alignment while clicking the picture of a shelf to assimilate analysis on product placement and the visibility of the products on the shelf.
  • Sales reps get customized classification of categories. This makes it easier for them to gather images corresponding to different classes.
  • Automated geo-tracking to find the nearby store previously entered by the field staff to curtail manual addition of the store each time sales reps go to gather data.
  • The deduplication algorithm detects fraud in regular audits of retail execution, thereby, significantly improving the data quality.
  • The data is secured and can be accessed by the client any time over ShelfWatch’s cloud.

Why ShelfWatch?

When it comes to achieving real-time image recognition, ShelfWatch plays the part perfectly. ShelfWatch is easy to integrate with any existing SFA tool. StreetSnap is successful in its niche, but if CPGs are already using any SFA app, then ShelfWatch can be easily integrated with the current model of SFA without compromising on any of the features. Setting up ShelfWatch is pretty simple. It only requires one image of the SKU for training. The algorithm is trained in such a way that automatically analyzes the pictures of the shelf and gives valuable insights like Share of Shelf, Planogram compliance.

ShelfWatch also provides a detailed Insights dashboard for the brands to monitor the latest trends. The dashboard provides competitive analysis by Counts, Presence, and Shelf area covered by SKUs and POSMs. It helps understand brand presence across all the stores along with geographical breakdown over a map overlay.

Liked the blog? Read our other blog to learn how brands are using image recognition and object detection technologies to achieve perfect retail execution.

Want to see how your own brand is performing on the shelves? Click here to schedule a free demo.

Author: Reashikaa Verma

Reashikaa Verma is the Content Marketing Manager at ParallelDots. When not blogging about AI in market research, her day is typically divided between reading, binge-watching, and petting dogs.

Leave a Reply

Your email address will not be published. Required fields are marked *