Over the last three years, I believe there has been roughly two decades of technological advancement. While that progress lets us tackle problems we couldn’t solve before, there’s still much to be done–especially in the retail space. One problem I have long watched is how we connect with the offline shopper in the same way we’re able to connect with online shoppers.
When customers shop online, we know so much about them. A login gives you an identity. Transaction history and items placed in a cart reveal interests. Abandoned cart behavior and removed items build an online persona.
From that persona, you can do what I call “persona-lization” – not full personalization, but a targeted, in-between approach.
Persona-lization Versus Personalization
Persona-lization recognizes patterns. If someone consistently buys diapers, you can assume they have young children. That inference lets you target marketing in a focused way. The online world gives us lagging and leading signals.
Why can’t offline do the same?
Today, offline retail mostly offers “lag measures.” After a customer or patient leaves, you know what they bought. You do not know what they picked up, tried on, or put in a cart and abandoned. Those missing signals hide why a shopper did not purchase. Was it price, fit, or options?
Technology and Feasibility
Finding those answers is the key, and it represents what technology may be able to help us with as we continue to advance.
Computer vision, deployed as an SDK (software development kit) on cameras, can tell you what a shopper is doing in a store. IoT tools—beacons, RFID (radio-frequency identification) and infrared sensors—can track item movement. Places like MIT’s retail innovation center are actively working on these problems.
The pieces exist, but they are not fully built. In eye care, you need an ecosystem of tools to track frames, count how many glasses are picked up and set down, and connect that behavior to purchases. The execution of this is difficult and layered, but once the data can be stitched together, it becomes powerful.
Short-term and long-term actions
There are short- and long-term plays to better understand your patients’ in-store behaviors.
In the short term, abandoned cart signals let you be tactical. If you have an online marketplace in your practice, you can test your patients with offers—15% off, 20% off—to see the conversion lift. What’s going to make them more likely to make a purchase? Then, you can make decisions based on data. Are you targeting value-driven shoppers, or those less price sensitive?
Long term, build a 360-degree view of your typical customer. Use “persona” data to recommend the next set of products. Then integrate that information with social media and ad algorithms to reach customers where they spend time.
For many ECPs, in-person purchases are driving most of the revenue. When we think about integrating technology, a practical in-store approach could be an app that ties login to location. If customers open an app when they’re in your optical, you can create prompts that automatically send them exclusive offers when they open it. You can also share directions to where specific frames are located in the store. Big retailers with strong apps, like Best Buy, are primed for this, whereas Home Depot is an example of where in-store navigation would help customers find what they need.
Implementation
In thinking about putting these strategies into practice, the first step is building the technology. Step two is staff training and maintenance. From my early startups in 2014, I know camera and computer vision can infer in-store personas. The missing ingredient has been integration across systems.
AI gives us the ability to recognize patterns at scale. What excites me now is using AI at the micro level, at the individual customer level. That lets you offer an even better, more complete consumer experience by connecting online signals to in-store behavior.
Read more features from Dr. Nafey here
Author
