the
best
customer
guide
Get our free weekly marketing briefings...
"They're always useful, and always brief"   :-)
Enter your email address to get started!
In the online world, it's not difficult to collect detailed data about customer behaviour because online shoppers always leave a digital trail, but capturing that kind of intelligence in bricks-and-mortar stores is much harder, according to Steve Jeffery, CEO for Brickstream. This article is copyright 2013 The Best Customer Guide.

On the internet, 'clickstream' data can be readily transformed into valuable insight that can be used to improve marketing strategies, pricing, product offerings and more. But nobody shopping in a store clicks on an in-store promotion, and Point-of-Sale (POS) data can only tell you about shoppers who actually bought something, not those that visited but left empty-handed.

In a world where the competition is just a mouse-click or screen tap away, retailers need to improve their strategies for capturing metrics about brick and mortar customer behaviour, so that they can better understand and manage this important sales channel. In fact, since the vast majority of all purchases (92% according to Gartner) still happen in physical stores, knowing how people behave in these environments is mission critical.

So where's the best place to start? Step one is getting an accurate count of how many people enter and are in your store at any given moment-all other key measures flow from this number. Here's why "people counting" is so important, and how retail managers can be sure they are getting the data needed to drive good decisions:

  1. Sales conversions: One Metric to Rule Them All
    POS transaction data tracks in store purchases and retailers often use this information as a measure of how well or how poorly a store is performing. But keeping tabs on sales alone does not give a complete picture of what's really going on. Conversion, or the ratio of people that buy to all shoppers in a store, is a key metric that tells a retailer how effective they are at actually "selling." A higher conversion rate means more people that shop are buying, and a low conversion rate means that while people spend time in the store, fewer of them are opening up their wallets. A low conversion rate is an indicator that staffing, marketing, pricing or the product mix may need to be adjusted in order to turn more browsers into buyers. This is where people counting (otherwise known as traffic intelligence) comes in. Stores need to know how many people have entered a store to calculate the conversion rate and obtain a more nuanced understanding of performance. For instance, a store that is considered a poor performer (according to POS data alone) may actually be doing a phenomenal job converting a low volume of shoppers. Without accurate traffic counts that show the store's high conversion ratio, the retailer may mistakenly punish an excellent store manager and miss addressing the real problem: finding a way to drive more consumers to that location to shop.
  2. Better Staffing, Service and Logistics
    As the old adage goes, you can't manage what you can't measure. Accurate traffic data can also be used by retailers to manage their workforce more effectively, both in real-time and over time. By reporting to store managers on a real-time dashboard the number of shoppers who are walking through the door throughout the day, managers can move staff to where they are most needed, whether it be to registers to ensure speedy check out or to the floor to assist shoppers. Historic traffic data can also be used to more accurately forecast future workforce needs, reducing the chance that stores will be left short staffed (especially important during the busy holiday shopping season) or waste money on overstaffing. Likewise, retailers can use traffic intelligence to analyse response to promotional campaigns, ensure health and safety compliance and even to figure out how often shelves need to be restocked.
  3. Addressing Data Capture Challenges
    Capturing data in brick-and-mortar locations requires installed in-store devices that are specifically designed to count people and track their actions as they shop. But even with this type of technology installed, separating relevant information from irrelevant noise can be challenging, as data inputs in the real world are far more variable and chaotic than what can be measured on a website, where every action leaves a digital artefact or 'breadcrumb'. For example, in an online store, every click is made by a known shopper visiting a specific webpage, while in a physical store, the technology must be able to discern between shoppers and objects like baby strollers, small children, or shopping carts. The environment can also play a huge role in the ability to collect accurate data in a physical space-bright sunlight, shadows, temperature fluctuations and other environmental conditions can interfere with a device's ability to discern what's happening. Blind spots caused by temporary/seasonal point of purchase displays or ceiling banners are another potential issue. If a data capture device is incorrectly positioned, or in a spot where it can be inadvertently or intentionally blocked, it may miss important details.
  4. Choosing the Right Technology
    Counting solutions available on the market today incorporate a diverse range of technologies that offer varying degrees of accuracy, so it's important to carefully test potential candidates under real world conditions. Some devices may work fine in low traffic environments, for instance, but lose the ability to count with precision when there are bigger groups of people. Others (thermal sensors, for example) may have trouble as indoor and outdoor temperatures fluctuate. Monocular video cameras capture flat images while stereo video technology collects three-dimensional information, which is more accurate under a wider set of variable conditions and can identify shopping units such as a couple or family shopping together. At a minimum, data capture technology should be able to: distinguish between adults and children, as well as between people and other non-human objects (shopping carts, for example); identify the direction of travel (are people entering or exiting?); and identify specific behaviours of interest (such as shoppers stopping at a promotional display). Beyond counting, solutions that are able to move beyond the front door to gather information about shopper's experience in the store (such as how long they wait in line, interactions with staff and movements about the store) provide rich insights into customer behaviour that offer very high levels of valuable intelligence.
  5. Staying Accurate Over Time
    Even the most sophisticated data capture technology can fail to deliver accurate intelligence if it's not properly installed, configured and validated. When purchasing a solution, retailers need to be sure they understand how a selected technology functions, carefully review configuration and validation procedures (and make sure that devices are deployed consistently across multiple locations), and become knowledgeable about the processes and tools available to support the proposed system throughout its life. Test data should be collected for several days to account for variations in environmental conditions, traffic volume and patterns, and additionally should be independently reviewed. Traffic accuracy should also be reported for 10-15 minute intervals throughout open store hours. Traffic accuracy of 95%+ should be the goal. Also keep in mind the important difference between counting individual people vs. counting shopping units. Depending on the type of store, families and couples can make up 20% or more of the traffic. But for accuracy's sake, a parent shopping with a child or a couple shopping together should be counted as a single "shopping unit" rather than as two shoppers. Allowing sufficient time to sort out these issues during installation will pay off in more accurate data over the long term.

"Today's brick and mortar retailers need to step up their analytics game to compete with online and mobile sales channels that are aided by a wealth of clickstream data. They can do this by counting store traffic and taking care to capture accurate data about shopper behaviour. This 'behaviour intelligence' will ultimately lead to better, smarter and more profitable in store decisions," concluded Jeffery.