Where Retail meets Tech

July 7, 2017

Can machines create ideal retail execution today

Written by Ayoub Khammari

In a crowded retail space, with stores competing with not only brick-and-mortar chains, but also the massive online market, superb execution is more important than ever. To get it right, brands and retailers need an edge.

One of the sharpest is a technology-enhanced feedback loop that lets you plan your strategy, execute it, and get real-time insights to continually hone your approach.

Some of this is classic business sense. Planning and execution have always been important, of course. But it’s that third step — real-time insights — that has only now become fully available to brands and retailers through exciting new technology like Planorama.


Hardware and Software: A Beautiful Friendship

How does it work? With a fruitful convergence of chips and code.

Today, digital cameras easily deliver high-resolution images to the internet through smart mobile devices. Couple this with powerful image analysis software on the cloud and you’ve got a plug-and-play system for close monitoring of your retail shelf and cooler space. And that means fast turnarounds on honing your overall strategy and execution.

"Powerful" software refers to automatic product recognition from digital images using deep learning algorithms. This approach, continually under development at Planorama, employs a form of artificial intelligence that handles realistically imperfect images with ease.

Imagine your products photographed at an angle or partly in shadow — not a problem! Deep learning algorithms recognize products at a level of accuracy approaching the human visual system. This allows for recognizing products on the shelf and automatically measuring a variety of merchandising KPIs.

The shelves and coolers are getting in the act, too. We’re seeing more embedded and web-enabled cameras in product areas, providing constant connection and consistent, high-quality images.

The combination of hardware and software advances unburden in-store representatives from spending time measuring the shelf manually — all while increasing the quality of the data. This allows merchandisers to dedicate time to more value-adding activities, such as promotions and selling.

 

The Dreams of a Field Rep

Imagine the difference that could be made for employees if technology worked for them rather than against them.

Let’s take an example of how we picture the ideal day of a field representative. He wakes up in the morning and checks his tablet to access his daily program. The program is defined based on the latest real-time insights from his area of coverage. Insights are fueled by shelf KPIs, electronic point-of-sale data, and market intelligence — all essential information that he had never been able to access before.

The field rep heads to the first store on his list and checks his tablet to identify what activities he should conduct in that store. That tablet’s suggestion would correspond to the biggest opportunity in the store, and by the time he finishes his visit, his daily plan is adjusted based on the latest insights collected in the morning. Both his route plan and in-store activities are no longer static but are optimized in real time to maximize efficiency and effectiveness. 

That’s how we see perfect retail execution: dynamic rather than static. With the latest actionable insights at their fingertips, field reps could work more efficiently and more effectively than ever.

 

Leveraging Technology

Cameras and image recognition software are a winning combination for manufacturers and retailers, but can we leverage other technology to make it even better?

We think so. For example, at Planorama, we’re now testing in-store robots to 3D-scan shelves into high-resolution images. Imagine a little squadron of robots wheeling around the aisles at night, taking consistent photos of normal shelf space and new promotional displays and handing off their data to the cloud.

We also consider smart glasses an exciting opportunity. Take for example Spectacles from Snap (previously known as Snapchat), available on the market for just $130. They generate circular photos or video with the 115-degree field of view that people can see. That means in-store reality as seen from the shopper’s perspective will be just a fast network connection away.

More than just robotic scans, this will provide incredible insights about not only the shoppers’ missions, but also where they look at the shelves, how long they look, and any stocking problems. This recorded, time-stamped, analyzable, and first-person view of shopping may sound like a marketer’s dream, but it’s actually right around the corner.

On the software side, there are advances too. For example, at Planorama, we are constantly tweaking our insight factory and fine-tuning our algorithms to further increase the accuracy and the speed of the analysis. But we also aim at combining other data sources such as electronic point of sale to boost the effectiveness of the insights delivered to field teams and headquarters.

With new creative advances in hardware and improvements in what the deep learning approach can reveal about products and shoppers, we’re confident that reality will be here before you know it.

 request a demo


Topics: image recognition, retail execution, artificial intelligence, machine learning