This is the story of an AI-powered footfall analytics solution that has been recently tested on a live mobile network. The solution provides retailers with rich insights that can drive up customer engagement and sales.
With the persistent rise of online shopping and e-commerce, brick-and-mortar businesses are facing tremendous pressure to keep customers coming in. Retailers, mall operators, and franchisees have been very vocal about the need to optimize and streamline their sales operations, in order to keep up. This indeed calls for finding new innovative ways to optimize costs on one hand, while enhancing in-store customer experience on the other hand.
We believe the key to accomplish the aspirations of brick-and-mortar retailers today is via enriched footfall analytics. The depth and breadth of intelligence sought after surpasses what traditional solutions can provide. While cameras are great, they are undoubtedly invasive to privacy and could prove to be quite pricy at scale. As such, there is a dire need for “footfall analytics 2.0”, offering rich insights into customer behaviors, demographics, and profiles. It should provide retailers with the opportunity to tailor customer experience according to the need of each individual.
In this short blog, I am going to share with you how our startup, MOBiSENSE, is taking a stab at this challenge by harnessing what 4G/5G mobile networks have to offer. I will also highlight some exciting progress we’ve recently achieved: testing with a business partner on a live 4G/5G network.
How it works
Mobile networks are just everywhere posing themselves as a viable and rich source of data. 4G/5G mobile networks amass tons of data from cell towers and smart phones alike. The so-called “measurement reports” represent snapshots in time and space of the network’s status. MOBiSENSE’s technology harnesses the power of AI to develop insights about people’s mobility using these snapshots. The solution we have developed offers the following bits of intelligence about visitors of a retail space:
- Time spent inside specific areas of the shop (e.g. at the cashier, in the seating area, in the parking lot, etc.)
- Time since a customer last visited this or any other branch (customers are uniquely yet anonymously identified).
- Typical visit times (hour of the day, day of the week), and seasonality trends.
- Hourly customer counts spread over the whole week, i.e. heatmap of footfall traffic.
- Time a customer spends transitioning between sections of a shop or within the retail space.
When coupled with sales figures and staffing schedules, footfall analytics of that sort sets retailers on a trajectory towards cost optimization and enhanced customer engagement.
Field trials with STC and Huawei
Having demonstrated our solution in a small-scale “lab-controlled” setup, we approached some potential partners to take this forward. It was STC and Huawei who embraced our efforts and enthusiastically offered to sponsor our field trials.
According to Forbes, STC is among the top 44 digital companies in the world, the most powerful telecommunications company in the Middle East for 2021, and ranked 8th in the list of top companies in the region for the same year. STC has been a pioneer in exploring new business directions and has spotted early on the opportunity to monetize its own network data, typically treated as strictly operational data (i.e. used to optimize network performance).
To realize such a progressive direction in practice, STC called upon Huawei to provide the right support, resources, and access to data interfaces. We have been actually overwhelmed to receive genuine support from Huawei engineers and technologists from 4 different offices, namely: Hong Kong, Shenzhen, Dubai, and Riyadh. The testing ran on a live production network, thus offering evidence that MOBiSENSE’s solution can be integrated virtually with any Huawei infrastructure.
We have tested in 4 different locations on 4G and 5G networks. Using an innovative machine inference engine, our solution beats any indoor positioning solution out there in terms of resolution and accuracy (mind you, our solution is not even a positioning solution in the traditional sense of the word). Our solution can also tell the mobility mode of customers (walking, biking, scooter, or riding a car) with very high accuracy.
We’re very grateful to STC and Huawei for embracing this project. Indeed, we are also grateful to KAUST, the cradle from which all of this has originally started. Efforts are underway to test our footfall solutions with other wireless infrastructure vendors. Please drop us a word at info@MOBiSENSE.ai if you have any question or interested to learn more.