What Video Analytics Can Do
In-store video analytics is a heady topic in retail these days. Retailers find themselves caught between a rock and a hard place: on the one side, they
have a desperate need to understand what is happening in their stores, and not on some one-off or occasional basis, but with the same depth of understanding
that they can achieve with their online sites. The more sophisticated retailers get in turning online analytics into shopper insights, the more pathetic
the information that trickles out of stores appears to be.
On the other hand, privacy is a big concern, and with every new data breach, consumers get wearier and less patient with having their private information violated, whether payment information, contact information, or purchase history or worse, video of them when they weren’t really aware they were being recorded.
The need that drives retailers to understand what’s going on in stores is not going away, and in fact, the overall holiday results only underscore the issue: in general, store sales are flat or down, and online is clearly benefiting. Stores are too much of an investment and cost too much to operate to run them in the black hole of information that they increasingly appear to be in – especially compared to online. The store experience is suffering, and retailers need real information, not corporate myths about what they think is happening, in order to respond.
I had the opportunity to speak with Marc Schwarzbart, the Vice President of Merchandising for Lolli and Pops, a high-growth candy retailer based in San Francisco, who is in the process of rolling out video analytics across the chain. Lolli and Pops leverages Prism Skylabs’ video analytics, which takes a unique approach to balancing privacy vs. the need for analytics by removing the people from the videos.
Here are three big opportunities that Marc sees for in-store video analytics.
The Basics: “I can step into any one of my stores at any moment from
Marc reports that in his past life at Old Navy, his team was always reacting to one-off reports of issues. “An executive would say, ‘I was in this store on the East Coast and they had limited inventory’ but the inventory system would show they had plenty. I didn’t have that visual confirmation of where the inventory actually was or how the store looked. And now I have that.”
Now, if anyone has a concern about a Lolli and Pops store, Marc can look straight into a store and see the shelves right away. Did the store put the inventory in the right place? Did they merchandise it properly? Marc reports that it’s much easier to make sure that stores are doing what they are supposed to be doing. It also helps the merchandising team stay connected to stores by enabling them to see more stores more often.
Root Causes: “Why did my store perform poorly last weekend?”
While the basics are extremely helpful, for Lolli and Pops this is only the first step. Marc’s real objective is to get to root causes when things don’t go as expected: “If I just rely on transactional data, I miss the opportunity to get to a root cause. Now, I can look right away – does the store look good? Why or why not?”
A common discussion at the retailer shifts away from second guessing consumers to using analytics to identify the source of an issue. Marc recalled a situation when new products featured on a display near the front of the store weren’t selling as expected. “The kneejerk reaction would typically be that customers didn’t like the product or the price. I pulled up the camera view and saw that the product was in an area that the traffic flow was missing. I had the store move one table three feet to the left to shift the traffic flow (which I could see almost immediately) and within just a few days I could see the results in my sales figures.”
The reverse opportunity is just as good. One of retail’s oldest problems is finding a way to make every store perform as if it was the retailer’s best store. Lolli and Pops uses video analytics to help all stores improve. If a store does really well, Marc has the ability to take a picture of what works. “I can send it out to other stores and say, ‘Hey, this works great – do this right now.’ It really closes the gap from insight to action.”
One area where Marc wishes he could spend more time is on traffic path analysis. Through the company’s experiences with root cause analysis, the team has realized the importance of looking at traffic flow right away. Marc reports that when product is underperforming, the team used to focus on whether consumers liked the products or if they are priced right. “But now my first question is ‘Did people even see it?’ If two thousand people come into a store but only two hundred make it to the international candy section, I don’t have an international candy problem. I have a traffic and signage problem. And that makes the conversation much more productive.”
How Do Stores Feel About It?
Lolli and Pops provides stores with the same camera views that corporate has, though for stores the emphasis is on loss prevention and security, while corporate is primarily using the cameras for product insights.
A bigger challenge for Marc is educating the corporate team on the opportunity. Right now, the company has about 20 stores with cameras, split between two banners, Lolli and Pops and Candyopolis. But getting people to add one more activity to their already full plate is always a challenge. Marc focuses on using the company’s successes as a learning opportunity, so that team members begin to demand access to the insights that video analytics provides, rather than having Marc try to push it on them. He says, “All I have to do is show them things like how much of an impact moving one table three feet had on a store. Show them that video analytics are valuable to their objectives. And then they are converts.”
There is a flip side to the capability, and that is resisting the urge to micro-manage stores. Marc has had times when he has checked in on a store and found an employee who should be greeting not at his post. Or that a delivery truck driver has left a pallet in the middle of the store’s entrance. But even with this level of monitoring, he has found that it’s still a better experience than the alternative. “Our CEO can ‘walk’ through stores from his desk and make pages of notes which now get sent to the store manager. It gives stores an opportunity to take feedback and make changes. This is much more productive than walking into a store, seeing something go wrong live and in person, and having an immediate reaction to it.”
Lolli and Pops is a small chain today, but Marc doesn’t see rapid growth as an inhibitor to video analytics’ future with the company: “No matter how big we get, I will always be closer to stores than I would be without Prism. At eleven stores, I can be in all my stores every day. When we get to 100 stores, it will be impossible. But at the end of the day, no district manager will have more than 10-20 stores. They could easily be in all their stores every day. And I can always be in my 3 best stores and my 3 worst stores every day.”
It comes down to having good data. For Marc, it was about gaining access to the same kinds of insights that can be driven from online. “If I just know that sales are good or bad, there are a thousand things that could be driving that and I may never know the underlying cause. With video analytics, we can’t get to a one-for-one match for different kinds of online analytics, but we can get to a really valuable insights about how our stores are being shopped and how to improve our business.”