Bringing the e-commerce superpower to retail stores: Actionable data with AI
We all know that e-commerce has an advantage. We're told that online shopping is the future. What we're not told is how we can learn from e-commerce and bring their powers to the real world.
Online shopping and e-commerce has a Superpower. It's everywhere, and many of us use parts of it in our daily lives, but it's difficult to perfect to the point where it matters. That super power is actionable data. Since 2011, physical retail sales have seen a 7% drop, primarily due to e-commerce being able to adapt and change so quickly. e-commerce retailers know exactly how much product is available, exactly which products are being viewed, and their conversion rate which means they can make optimizations and actionable decisions instantly with the data they collect.
The e-commerce superpower vs physical stores
You're probably thinking that you're already using data. After all, its such a universal concept. Why then, does e-commerce have a data superpower? Let's look at e-commerce's approach to data, and how our physical stores match up.
1. Superspeed: digital data is built to be collected from the ground up
The digital renaissance has provided us with huge swaths of easily collectable raw data. Every action we take online is already in numbers. Every mouse click is translatable to a string of coordinates, every page we go to has its own entry, and every time we go online we have a trackable persona. There isn't a single thing we can't understand, because the data is already there in numbers.
Stores: interpreting the real world into numbers poorly
Making the exact type of optimizations that e-commerce retailers do online in our physical stores just won't be possible. Real life doesn't consist of strings of numbers, it's real people doing real actions. We're working with an infinite number of variables, and only 5 senses to work with. We have to find a way of interpreting the actions we want to understand as numbers. Sure, we can have real people manually counting or scouting stores, performing surveys, or using decades old technology of people counting. The problem is, these are shortcuts that take forever to process and can only interpret vague ideas.
Could you imagine if Amazon had to assign someone to manually count every single person that browses their website? If e-commerce is able to natively understand so much about their websites, why can't we do the same with our physical stores? We need to better interpret the real world into smarter data.
2. Superintelligence: E-commerce creates data that drives actual decisions
If you've ever had to comb through thousands of plaintext entries, you know how too much raw data can be just as useless as too little. E-commerce's digital advantage isn't just in collecting the data, but it's ability to capture and automatically process data that's easily understood to make better decisions. We're talking website heat maps, customer journey maps of individual customers, and a complete understanding of every action every user takes on their storefront.
Stores: playing catchup on data interpretations
Our physical stores on the other hand are playing catchup. We're not collecting enough useful data to make confident decisions, nor is it easy to automate our systems as we're locked into the real world. We need sensors that enable us to automatically build a dynamic understanding of our real world, may it be with new technologies or processes, we need to be leaders, not playing catchup. For example, heat mapping users has become a web commodity that's been making its way into retail. Serbian fashion retailer Legend World Wide deployed computer vision on cameras and sensors in stores to track customer journeys and understand how products move around their locations. They were able to quickly identify issues with store layouts that lead to many men leaving the store quickly after entering due to skimming through women's clothes first when first entering. A solution was developed to add navigation signs indicating the direction of the male section, solving the problem. Computer vision not only enabled Legend to save the time an analyst would have needed to come to the same conclusion, but make faster and more confident decisions to rapidly save money & time.
After all, if online shopping has an excellent understanding of how their users shop and can gain so many valuable insights, why can't we do the same with our physical stores?
3. Shapeshifting: instantly testing changes to optimize every small detail
Having instant access to data reveals e-commerce's 3rd superpower: being able to test changes on the fly. Websites are able to use every single customer as a experimental subject for A/B tests, which means they can fail fast and adopt the most effective changes from targeted content, website pathing optimizations, product listings, and more. E-commerce is able to experiment on new layouts for their storefront and optimal text for product listings to the smallest of changes like the color of a button or shape of the checkout box for virtually no cost, instantly, and discover results that you would never expect. For example through rapid A/B testing, software platform WorkZone increased its leads through its testimonials page by 34% by simply changing the color of the logos on their testimonial form.
Stores: stuck testing changes blindly
While online stores can have data and test changes instantly, it either takes our physical stores months of research to gain a slight understanding of what goes on in our stores... if any at all before it becomes outdated. The fastest widespread form of data collection, people counters, turns our stores into a black box we can't look inside. All you know is that a person entered the box and what they take with them when they leave, but nothing about what’s happening inside to lead to that decision Could you imagine if the only information Amazon had to work with was how many people entered the front page, and how much stock they have left? We're lacking data, and having people counting as our only sensor just doesn't make sense when we can collect so much data from online shopping to make powerful decisions. From assisting data analysts by making their job easier with actionable data, to simpler, faster decision making with footfall analysis and heatmaps, we need to unveil the black box of mystery that is our individual stores, and understand everything that's happening inside of them. Why can't we test as fast as e-commerce and make more confident decisions?
4. Mind-reading: online enables simple services for better customer experiences
From building better customer journeys, better targeting products, to understanding customer pain points, e-commerce is building build better customer experiences with minimal effort. When you're shopping online, online stores are building an adapted customer experience that's completely unique to each customer.
Stores: we're missing out on sales by not prioritizing experience
Our physical stores has a strength that we need to capture: it's ability to deliver unmatched customer experiences and convenience. When a customer has the choice between shopping on their couch vs doing the same action in a boring store, the choice is obvious. If retail can better understand customer experiences, retail becomes a unique experience in itself. From better understanding where customers need help, providing line estimates, or a museum like experience, creating unique experiences that only physical stores can provide is more important than before. Why aren't we able to automate customer experiences physically as well as we can online?
5. Superhealing: digital technology automatically improves over time
Finally, the last part of e-commerce's superpower is it's digital ability to automatically improve over time. Not only is it simple to test and deploy new technologies to stay ahead, but the technology is inherently capable of learning over time. From AI ad delivery optimizations, to trend analysis, the more data e-commerce collects, the better it's algorithms learn to collect it.
Stores: we take forever to adapt and upgrade with hardware.
Our physical stores on the other hand are forced to upgrade our hardware every time we want to improve. If we want to have better sensing technology, we're forced to manually switch out hardware. If we want to adapt and upgrade our technology automatically, we need to switch to software and merge the digital world with the physical. For example, Target has been exploring computer vision to make better decisions and improve their customer experience. They've deployed computer vision in over 1,500 stores to do everything from advanced people counting, to planned developments like shelf management and store management. All of which now can grow in effectiveness automatically, and be easily upgraded with a simple software update.
Why's it so hard to do better?
1. The retail environment is complex
Something that works somewhere doesn't work elsewhere. From clothing stores, bookstores, grocery stores, home improvement centers, and department stores, you can't simply apply something that works at one type of retail environment into another. New technologies are complicated to adapt and deploy broadly.
2. Massive infrastructure investments take forever
Implementing new dedicated hardware is difficult and as a result the process takes forever. You never want to make a mistake but in the time that hardware takes to develop, it will be outdated by the time it comes out. In an ideal world, you'd be able to use your existing technology like your security cameras as multi-purpose tools to speed up investments and improve over time.
3. Any projects you undergo are expensive and risky
New infrastructure is always expensive, and no one ever wants to be the first to take the plunge. You're spending an incredible amount of money on hardware that will be out of date, possibly leaving you stuck with devices built for an outdated purpose.
4. It's difficult to adapt and change things once they're set up
Setting up anything permanent means any changes you make to your stores basically means the entire process starts over. That just doesn't make sense when you need to test changes quickly if your sensors need to take forever to be adapted every time.
It's time to get ahead
E-commerce's superpower isn't difficult to understand, we just need to understand our roadblocks and use tools that are more adaptable, less risky, and that enable us to combine the e-commerce superpower with our retail strengths.
It's time to make better decisions in our brick and mortar stores.