Like most analytics and software tools, the biggest stumbling block for people-measurement platforms like our DM1 is translating tool into business use. That’s often a particular challenge for analytics – which is why the gradual transition of digital analytics from ROI based sells to table-stakes based sells was so important. Measurement systems may and should have good ROI, but proving that ROI is next to impossible since it’s by definition a function of what the analytics finds and whether those finds are acted on.

When it comes to alerting systems, the case is not quite so troublesome. This overview of alerting use-cases isn’t going to be remotely exhaustive – we’re constantly surprised by the use-cases that people bring to us. But this post should give you a rough idea of the types of use cases that are available so that you can apply that knowledge to your own business to see if anything fits.

We’ll carve up alerting use-cases into three main buckets: operational management, customer journey, and security and loss prevention.

Operational Management

Operational Management use-cases form the largest and probably the most common set of alerting use-cases. We’ve posted often about queue management, and line-based alerting is one of the most common applications of people-measurement. You can alert based on line length, wait times, predicted wait times, and excess processing capacity. Whether in airports, theme parks, hotels, or retail cash-wraps, lines are a potentially devastating part of the customer experience and alerting helps ensure you deliver the best performance possible.

Operational management based on people measurement is hardly limited to lines. You can use alerting to manage labor allocations – alerting based on occupancy or staff to shopper ratios. For mid-sized to large stores, dynamically moving Associates to slammed areas just makes sense. Any place and any application where you have the opportunity to more efficiently move staff to need, people-measurement alerting can work. Nor is it simply a function of crowds. We’ve seen real interest in things like maintenance scheduling. How often does a restroom need to be refreshed? It depends on usage. And that usage can be measured – allowing you to dynamically allocate facility staff to the places most in need of service.

Another form of operational alerting is as a trigger for changes in operational strategy. If, for example, you add an employee to work a drive-through line or, at a retail bank, to work the teller line, depending on the length of the line, you can use alerting to trigger the operational change. This kind of alerting can ensure that you implement and remove strategies quickly – ensuring optimal labor usage even when on-site managers get distracted.

You can also use operational alerts as a way to crowd-source public information. Letting people know which lines are least crowded. How long they’ll probably have to wait for service. What rides are best right now. Even where to find an employee or which reserved Conference rooms are actually occupied.

Core capabilities for operational alerting are person-type identification – particularly staff vs. customer – accurate occupancy measurement, and alerting control. The first two are self-explanatory, but the third needs a brief explanation. No alerting system can afford to be too noisy. We’ve all had the experience of reporting or alerting systems that inundate us with notifications; inevitably, noisy system result in rapid tune-out. That means a key capability for any operational alerting system is the ability to track recency of alerts and throttle their pace appropriately.

We haven’t even considered back-office and process related opportunities for alerting – operational management just encompasses so many potential use-cases (a lot of which we’re probably not even aware of). But if you sometimes need to move people around in a complex space, there’s a pretty good change there’s an application for people-measurement based operational alerting.

Customer Journey Optimization

Though Operational use-cases form the back-bone of real-time people measurement alerting functionality, the use-cases around customer journey are probably the most fun and interesting. There are so many different ways that this kind of alerting can be implemented – and many bring a truly novel element to the customer journey.

The most obvious customer journey alerts are designed to direct Associates to help shoppers who may need assistance. These are usually dwell-based alerts – looking at the time a shopper has spent in the store or, more commonly, in a particular area. Because people-measurement is so precise, this can literally be at the individual display level. If a customer is standing in front of a locked display (and needs their NyQuil), a good people-measurement system can detect that. If a shopper has spent five minutes looking at golf clubs, sending an Associate to help might make sense.

Generalizing this, customer journey alerts are often used to send help to customers when they are in an area that always requires it. The locked display is an example. But if you have a customer at any unmanned support area (from a pickup station to a customer support desk to a ticketing booth to a cashwrap), then getting them the help they need as quickly as possible is essential to good customer experience. In general, alerts for this kind of behavior are dwell based, but are often very short. A customer shouldn’t have to stand at a locked display for very long before help is dispatched.

The golf-club case represents the other common customer-journey use-case. Targeting sales assistance to high-value customer behaviors. High-value items in a store are, obviously, bigger shopping decisions and cases where sales assistance is most likely to be important. We don’t usually help picking out a pair of socks or a paperback book. But if we’re buying a bike or golf equipment or electronics or jewelry, getting sales assistance ranges from beneficial to essential. In fact, stores are often setup so that you can’t buy these items without sales assistance. Footwear inventory may all be in the back-room. The TV you buy isn’t the one on the floor.

The overall idea behind this kind of alerting is simple. Send Associates to help customers where that help is most needed or most impactful. Typically, that means looking for dwell-times in areas of high-value.

Note that you can flip this strategy around and drive it from the Associate perspective. Instead of sending alerts based on individual shoppers, you can direct an Associate based on the highest-value opportunities in the store. In other words, each Associate is told who to help next when they finish an interaction. This means scoring every shopper in the store and picking the one that has the highest predicted impact from an interaction. Building this kind of dynamic allocation system goes beyond alerting, but it’s essentially just the same capability on steroids.

Another use-case for customer journey alerting is identifying customers that have a problem. They can’t find something. A machine is broken. Something about their journey is off. This kind of alerting can really improve customer experience in a facility and is most definitely not limited to retail. In a store, you might look for a customer walking up and down a small aisle area – strong behavioral evidence that they can’t find something. But in a train station, you might look for someone lingering at a ticketing machine or at a route map. The ability to target interventions to customers who may be struggling transforms the nature of the physical experience.

Of course, this assumes that there are behavioral cues that show a customer is struggling. That’s obviously not always the case. Just keep in mind that what’s getting measured is a customer journey. Everywhere they were and spent time. Wherever that path gives you real cues about potential customer problems or opportunities – there’s a use-case for customer journey alerting.

Probably the coolest and most interesting use of customer journey alerting is in cases where you can use what a customer has done to tune the experience you’re getting right now. If your customers are interacting with displays or experiences and you could use prior knowledge of that interaction to tune the next one, you can use id-based alerting to let each experience know what else the customer has done. This kind of alerting can make experience in a physical space as personalized as it is in the digital realm.

Loss Prevention and Security

The third main use-case for people-measurement alerting is loss prevention and security. To be clear, loss prevention and security aren’t really something we focus on. Digital Mortar has always been about operations and marketing. And believe us, security is a very distinct beast. Not only is the use-case different, the stakeholders are different and often the technology is different as well. So we don’t normally crossover into security applications.

But there’s two aspects of this that are worth calling out. First, there are times when it’s possible to dual purpose the underlying sensor technology to provide both crowd analytics and security. When you can do that, you lessen your overall hardware investment and you significantly improve the ROI on your sensor purchases.

And second, and this is where our current thread comes in, people-measurement alerting can sometimes span both applications. When you have a generalized alerting system based on how people move, you aren’t limited to operational and marketing use-cases.

Still, to be clear, the kind of measurement we do is about people’s journey. We don’t (and can’t) measure whether someone just pocketed an item. All we know is how people moved.

What that means is that people-measurement alerting is most useful for things like perimeter monitoring. LiDAR, in particular, is stellar for this kind of dual purpose functionality. For example, we have a gas-station/convenience store client that uses our system to measure the full customer-journey from pump to market and throughout the market. On the outside, we use LiDAR to measure parking space usage and gas, understand how often customers come to the market and don’t get gas, and how often the parking spaces are full. On the inside, we measure how people use the market and how if differs depending on the type of customer journey. All cool stuff. But since LiDAR is measuring the outside, we can setup alerts that do perimeter monitoring during the off-hours. LiDAR works without lights and is fantastic for this kind of security application.

You wouldn’t put in a system like this just to get after-hours perimeter alerting. But since the cost is zero once you’ve installed the analytic system, it’s just another way to get better ROI.

The kinds of alerting applications you can support include this kind of perimeter monitoring as well as more specialized applications – things like entering restricted areas. Emergency occupancy reporting. Occupancy exceeded alerting. Directional violations (like people going the wrong way through an airport exit). Even Velocity outliers (fast movers in the space). If it just needs people journey measurement, chances are it can be used for an alert!

And while t security has never been our focus, getting better ROI always has been.

For a short overview of DM1’s Real-time alerting system, check this out: