Is That a Visionary Product Display or a Product Display With Vision?
When the conversation turns to the latest in-store tech trends, it usually revolves around tablets for sales staff, location tracking for consumers or digital signage. While that is "technology," in the grander scheme, it’s not all that high tech. However, there is a new genre of retail technology that is transitioning from research to reality and it really does have potential to change the dynamics of what we’ve come to accept as shopping.
Call it product displays with vision. The emergence of "smart shelves" — store displays that can see (detect/sense) shoppers — has begun. These units utilize 3D depth sensing cameras (like MS Kinect) to "watch" consumers as they browse for products and make shopping decisions. Combining a number of features that include body positioning, gesture recognition, gaze tracking, facial expression, age/sex analysis, and voice recognition, the potential sophistication and power these systems can have is unprecedented.
Currently, indoor consumer location tracking (primarily via mobile devices) is still in its infancy, leaving much to be desired. Depending on the method used, mobile device tracking may be incapable of knowing precisely where a shopper is standing, and almost certainly can’t tell which way they are facing, what they are looking at, or what they are doing. But, through the use of high-res cameras, depth-sensing technology can determine those things, positioning it to leapfrog way beyond rudimentary location awareness.
In contrast to just knowing that a consumer is near the shampoo section and sending them a mobile coupon or playing a commercial on a digital sign, true behavioral analysis can provide detailed real-time information and response, marketing analysis, and display/planogram feedback.
The technology is being scheduled for launch by Mondelez International (Kraft, Nabisco, Cadbury, etc.) in 2015. According to ABC News, the company’s stated goal is to "understand how shoppers see, scan, spot, show interest and select products from the shelf in the store."
As is often the case with new consumer technologies, invasiveness and privacy questions come to the fore, so time will tell if brands are willing to risk consumer backlash in the hopes of enhanced shopper marketing opportunities.
It is expected that in most instances this type of "tracking" will be implemented so that shopper data is anonymous; however, it’s possible to identify consumers either with or without their permission.
Discussion Questions
What benefits do you see coming for retailers and brands from “smart shelf” technology? What limits and assurances will be needed to gain consumer acceptance? Do you see significant advantages over other emerging shopper tracking technologies?
Smart shelf tech lets retailers get much more granular data about how products are shopped. It can help retailers address out-of-stocks (still one of the most expensive problems in retail) and determine unfulfilled demand. It helps merchants replace decades of intuition and urban legend about the most effective adjacencies and merchandising techniques with actual shopper data.
A shelf that identifies you by age? Just what the world of retail was looking for!
Seriously, the degree to which one sees this kind of technology as useful depends on the degree to which one believes retail is more science than art. I have maintained for years that retailing’s problem is less with the amount of data gathered – or potentially gathered – and more with the lack of tools for translating that data into insights. This may threaten to be more of the same.
I think that technologists need to go back to the “Hawthorn Effect” studies of the early 1950s which demonstrated (as have physicists) that observation changes behavior. In this case customers who know they are being “watched” by – say a shelf – may in fact start to shop differently.
So, what is the real value of the data captured by the intelligent agent? It’s a question that will have to be resolved over time.
As to privacy, I wouldn’t want to be the first company in America to let shoppers know I was recording their age, gender and every movement. Not until people get over the shock of the N.S.A. listening to all their phone calls. It’s just a tad too Big Brotherish for the times.
I am not clear how the benefit is going to outweigh the cost. Also, as a consumer I would avoid stores that had this technology. I don’t feel comfortable having cameras watch me while I shop. Especially if they can identify me. The FBI and CIA would love this. I would encourage retailers and brands to take baby steps. The backlash could be huge and have long-term impact.
Building customer trust is 101 retail. Does this solution accomplish that?
I am all for stores doing more in-store tracking, but it has to be done with the consumers’ permission. What is a store giving the consumer “to understand how shoppers see, scan, spot, show interest and select products from the shelf in the store”? If a retailer/brand doesn’t get the consumer’s permission, I think the penetration rates will be low.
As far as I know, Prism Skylabs, is the current state of the art in store monitoring and while I don’t know exactly how it works, the following statement using “privacy-protected” indicates they understand the consumer’s concern, “Prism Skylabs condenses customer interaction into stunning, privacy-protected imagery and reports.”
Smart shelf technology and the ability to track consumers through mobile phones is getting very sophisticated. There are many benefits to be gained by retailers IF they can leverage this data to create meaningful intelligence.
But for those retailers struggling with harvesting loyalty programs, these new technologies produces BIG DATA that few are equipped to incorporate or even process.
The benefits for the consumer are less clear. Just how much privacy are you willing to give up to a get perceived benefit? At the end of the day it is the consumer that votes with their wallet where they shop.
I’m not sure this has any consumer benefit and wouldn’t be surprised to see a lot of pushback from customers. Seems to me that retailers who claim to be trying to induce healthy eating are going to have difficulty reconciling this with a shelf that says, “wouldn’t you really rather have an Oreo?”
The approach will not be successful if it is about what the technology can do. It will only be successful if approached from the perspective of what it can do for the consumer. For example, if the technology collects my age at the shelf, what does that do for the customer? The products will not change. Would it really be more persuasive to then receive a message related to my age about Cheerios? What is the point of gathering my age except that the company has more information? If that is the value consumers will not interact with the technology and could be offended because private information is being collected and stored for not purpose for them.
Just because technology makes something possible does not make it a desirable activity. If consumers are going past 1600 items a minute in a store, do they really want to take time to stop and interact with devices (after the novelty wears off)? Does it really provide value to the consumer?
Who will deploy and “own” the smart shelf implementations? If retailers own the infrastructure, what kind of data and control will be shared with the brands? It’s clear the brands are looking for ways to get more data and the ability to engage the shopper directly. I’m sure the data gathered can be very useful for both parties, but will there be a framework set up to foster collaboration?
As for consumer acceptance, there is much that could be done anonymously that would provide good analytics and insights into behavior. However, once the brand (or retailer) wants to begin communicating with a shopper with personalized offers and engagement, the game gets much more complicated. Most likely the anonymous levers will be pulled first, while the strategies and tactics for personal engagement are phased in with careful steps. And rightly so.
Monitoring shopper behavior, movement and traits is all well and good – whether the method is Kinect detectors, overhead video, mobile device tracking, eye-movement tracking, electric eyes, pressure-sensitive flooring, or white-coated technicians with clipboards and stopwatches.
This is a lucrative business area for market researchers and agencies because the collected data lends itself to sophisticated analyses and visualizations, like heat maps, animations and time series. Brands are desperate for insights that can deliver an in-store performance edge and justify marketing plans, and they appear willing to spend to obtain them.
The “smart shelf” may hold an alternative meaning, however, one that is less about monitoring shopper behavior and more about monitoring what’s actually on the shelves or racks at the moments of truth. Utterly reliable merchandising implementation is a prerequisite for any accurate analysis of shopper response. Product pushers, weight-sensitive shelving, and various store scanning systems are potential solution elements.
Shoppers can’t buy an item that’s out of stock. They can’t consider a promotion that’s not on display. They can’t respond to a schematic change that is not implemented. They can’t try a new item that has not been cut in to the shelf.
I don’t worry too much about shoppers’ skittishness over in-store privacy in connection with shelves that stare back. I think most realize we are going to watch them no matter what. If the right products are available, we’ll get a fairly true read of their preferences. If the merchandising is sub-standard, no amount of smart shelf wizardry will help.
Okay, so this type of technology allows the retailer to do a sort of curating for each customer and we all know from the the likes of Pinterest, et al, that consumers like curating.
The question remains then, will this curating result in purchases? The consumer won’t mind so much that you recognize them as male/female, younger/older as long as their complete identity is not being collected and there is something for them.
If the result is purchases and this exceeds the cost, then retailers may buy in. Do I see this as wide spread? Not for now. Retailers have so much data piled up that they aren’t acting on, it will take a special category, campaign,etc to motivate them to move on this. IMHO
Benefits include opportunities to engage with the shopper in novel ways. The technology should be used to “give” something of value to the consumer while “taking away” the data of value to the retailer or brand. Down-level of this shopper tracking IT, there is still value to be optimized in QR codes and near field communication to their full potential, which requires orchestration between merchandising, sourcing, marketing and store management.
Benefits include opportunities to engage with the shopper in novel ways. The technology should be used to “give” something of value to the consumer while “taking away” the data of value to the retailer or brand. Down-level of this shopper tracking IT, there is still value to be optimized in QR codes and near field communication to their full potential, which requires orchestration between merchandising, sourcing, marketing and store management.
There is a lot that a “smart shelf” can help with, without getting into what is considered “personal space” of a customer. Retailers and CPG companies will find use for a “smart shelf” in more traditional supply/value chain activities. The data that such a smart shelf will provide can be used to determine inventory levels, to automatically trigger replenishments, to figure out which products are being looked at more often (better packaging?), which products get picked from the shelf, whether the products are being positioned at the correct on the right shelf etc.
While “smart shelf” technology may be ready to also perform deep behavioral analysis, as pointed out in the article, Retailers will find that their customers wouldn’t be too willing to play ball (yet) and are better off using it for more traditional benefits.
I have a few observations and comments to add to this discussion:
1. Video technology has been used at retail for many years such as queue length management, customer service management, shrink, security or out of stocks. The benefits to retailers are obvious now and we take much of this for granted.
2. In-store tracking has been performed for years too. Historical approaches to do it have not been perfect: e.g. having a researcher on your shoulder, eye-tracking or shopping in a “lab” store.
3. As technology improves, the ability to bridge and improve historically disparate solutions becomes a natural choice for retailers to make.
4. Online: it is already possible to track every click and page view. By linking this data to standard demographic data, or even just using clever algorithms, this can already include age and gender and lifestyle and so on too. Outputs from this research can certainly improve retail choices like pricing, promotions, assortments, recommendations and substitutions already. Many customers seem to like these benefits.
5. Tracking this level of detail in a physical store is somewhat different. The outputs and benefits may be similar, but the methods and perceptions of this are not … yet?
6. Differentiating the retail experience whether online or in-store (and ideally “omnichannel”) with differentiated or even personal prices, promotions, recommendations, offers, substitutions, recipes is very compelling for many customers and therefore retailers and brands.
7. It can’t and won’t be a one-size-fits-all. Some customers will want more or less of this, likewise retailers and brands will play more or less in this space too (some by design, some due to ignorance!). Customer choices will continue to drive the dynamics and ultimately the retail successes and failures.
8. Some of the other interesting dynamics yet to unfold in all of this are (a) who will “own” and control the data in future, and (b) what happens when products or objects can interact and communicate with other products or objects directly? Will they be tracking us? There’s a scary thought for Halloween week!
Interesting topic….
This technology has been evolving for literally years at this point. I commend innovative companies as described in the article for leading the way to better understand shopper behavior. Within 18 months or so we will begin to see some proliferation of adoption of a couple effective tools in the global marketplace by those CPG brands and retailer partners that choose to leverage the insights available to them via these technologies.
Once these tools are proven, I see huge benefits for those that employ them.
This technology is still far from implementation. Being able to assure consumers that their information is both private and protected, once the technology has “caught up” to the level where it can accurately determine a consumer as they walk down a store aisle, will become one of the largest obstacles to overcome.
Right now this is the kind of “Gee Whiz” technology which excites marketers eager to try the next big thing. I should know – I’m one of them!
Even so, beyond the (inevitable) consumer backlash is the issue of actual effectiveness. Initially at least, these kinds of programs need to iron out wrinkles before they are truly effective for marketers and accepted by shoppers.
Technology seems to move forward in increments; its reach exceeds its grasp, until everyone catches up.
Ultimately, it is probably a bigger step forward to recognize the device in the palm than the face at the shelf.
Tracking the true effectiveness of a planogram has always been a challenge. Though sales is the measure which most use, its difficult to account it to only planogram changes, since so many other factors impact sales. Moreover, the complexity increases when the item is placed in multiple locations. Technologies such as these help to track the time spent by the user on the aisle, what has been picked up, what has been kept back, and others. Though a cctv camera could also capture a lot of this information, it is difficult to extract. Most of these technologies need to mature little more for commercial scale and usage.