Share this post:
When you next go shopping, glance at the shelves around you: the options are endless. Today’s consumers face so many choices that it can be hard for retailers and consumer packaged goods (CPG) companies to keep track of their products.
The challenge for these stakeholders is ensuring that popular goods are always in stock and positioned where customers are most likely to buy them. The ability to make on-the-spot stock decisions dictates which companies stay ahead of the curve.
Unlocking the mystery of store shelves
At Intelligence Retail, we wanted to provide fast, accurate product information to help retailers and CPG producers achieve competitive advantage. We developed a solution that uses visual recognition technology to analyze photographs of store shelves. These photos from smartphones and tablets are used to identify products, and machine learning algorithms calculate key performance indicators (KPIs) around presence, position, promotion and pricing.
To ensure that the solution could provide useful results fast, we needed a powerful infrastructure platform and an intuitive user interface. Our existing cloud platform simply wasn’t up to the task, so we began our search for an alternative.
Ramping up performance
Intelligence Retail migrated its solutions to IBM Cloud bare metal infrastructure, featuring GPU servers designed to handle compute-intensive workloads. Our testing found that IBM GPU Cloud servers outperformed the competition by 40 percent. Working with IBM gives us access to the latest technology developments at a very reasonable price. We also gain expert advice on developing our offering.
With leading-edge GPU computing supporting our solutions, Intelligence Retail can offer users very short response times. Complex photographs of store shelves are analyzed in no more than 10 seconds and have an accuracy level of 95 percent.
To present insights to clients, we deployed IBM Cognos Analytics. It powers customizable dashboards for sales performance and price, so users can drill down into KPIs to discover their best-selling products and strategies. They can aggregate KPIs by region, retailer, store, brand or visitor.
Winning the retail game
Up and running in the IBM Cloud, we’re ready to take Intelligence Retail global. With rapidly delivered insights from Intelligence Retail, retailers and CPG producers can increase sales by two to three percent and cut audit costs by up to 80 percent.
Two years into our partnership with IBM, we’ve extended our reach across Europe, and greater expansion is on the horizon. Thanks to the scalability afforded by IBM Cloud solutions, we can serve companies of any size at short notice.
Next, we plan to explore the demand forecasting capabilities of IBM Watson. We want to build on our solutions’ insights to provide advice on what customers should do next. By giving retailers and CPG producers another tool to get ahead of the competition, we’ll stand out in the market even more.
Read the case study for more details.