As we move into the future, we’re going to see AI-driven technologies like AI chatbots replace workers in a variety of roles – retail will be especially affected as we are beginning to learn that many processes can be more effectively handled by a machine than a human. Just as many manufacturing facilities are already using “smart” machines to precision build items, commerce will undergo some significant changes, thanks to new technology.
Certain sects of sales, customer service, and other areas within the realm of exchanging goods and services are already using ai chatbots to improve customer engagement. In fact, the app we just built for inHouse uses this technology to help customers with bookings and other services. Now is a good time to explore this technology to understand why we’re moving in this direction, what the implications are for the market, and how we expect ai chatbots to actually save money for retailers.
Why human labor is going to the wayside
As a species, humans have accomplished remarkable feats. Perhaps our most impressive accomplishments relate to things we’ve built from vehicles that can transport us to devices that enable us to communicate anywhere in the world in an instant. But despite our abilities to solve problems and build solutions, we’re still not perfect at using that which we create because of the fact that, well, we’re human.
We become distracted or tired and crash vehicles. We sometimes miss details when working on a project, resulting in a faulty final product. In some cases, our emotions cause us to deviate in our actions.
Machines we’re building to take the reigns over something that typically a human would control essentially circumvents classic human conditions. A self-driving truck can’t get tired and veer into traffic. An AI chatbot can’t get annoyed with a customer and tell them off. A machine won’t yearn to be with its family back home during a long work week.
While we’re able to build and use a myriad of products we’re just not that precise. With the ability to “think” and correlate information input into a system – whether it’s a message from a user or data collected by a sensor on a device – machines and apps can more accurately make decisions and take appropriate actions than the majority of humans.
In Andrew Yang’s book The War on Normal People, he jumps right into how automation will cause something he calls “The Great Displacement” where workers in a few key industries (later discussed in chapter 4) will eventually be replaced by AI. We need to be prepared to learn how to most effectively apply this technology as well as mitigate the consequences of moving in this direction as the market does not ignore more efficient means of production and distribution.
Cost savings in using AI and chatbots for retail
Many major retailers have already implemented AI to accomplish various tasks in a novel fashion. The article linked in the previous sentence describes some great adaptations of AI, such as the LoweBot at Lowe’s that helps customers locate items and manages inventory with computer vision.
Scanning shelves and monitoring inventory feels like something that could be accomplished with traditional programming. But the reality is that AI does more with its computations than running queries to return a list of results.
The LoweBot can handle more advanced processes like helping customers when they’re not sure exactly what they’re looking for. Not all of us are savvy carpenters so there are times when we know, say, the function of an item or tool we’re looking for but not sure what to call it, much less how it fits into the organization of a large warehouse. By using context, the LoweBot can return more dynamic results than running a query that would simply pull items that contain matching keywords from an inventory database.
For example, maybe you’re doing a DIY home improvement project by running a custom length Ethernet cable in your house and know you to affix “an end” to it so that it can plug into a device. With proper “training,” the software should be able to determine that you either need crimpers and RJ45 connectors or if you’re making wall installation, you’ll need an Ethernet wall plate and a punch down tool.
Tools such as the LoweBot reduce operational costs in a couple of key ways. In the example above, this implies that this task would be alleviated from human intervention, thus allowing workers to tackle more complex tasks. The bot also keeps track of inventory while not engaging with customers, hence ensuring that products are stocked at all times, even after the store closes down. Unlike workers that might have a lot of knowledge in one area (i.e. installing cabinets, electrical, landscaping, etc.) the bot can specialize in multiple areas by answering basic questions and keeping track of inventory, both of which reduce expenditures on labor.
In other businesses, the AI chatbot exists as software that handles customer service tasks that would typically require human interaction. The customer engagement company Com100 details some of the harder figures for actual savings. One great example with calling or messaging centers – by using an AI to field messages, the cost per interaction drops dramatically as it allows customer issues to be resolved quicker, with more accuracy, and less overhead as it’s typically cheaper by about 33%, on average, to make outbound calls should a customer request a call to speak with a human.
A good example of this is the Amazon chatbot used for customer service. While the neural-network-based system is still being evaluated, it’s being applied in customer-facing, production systems to solve problems – most customer issues are resolved through the AI which is measured by the number of customers who need to call back in the future to resolve an issue. For the most part, the AI can guess the issue by the user behavior which brings up a multiple-choice set of options allowing the user to get started resolving their issue.
We built an AI chatbot named Pepper into the app we built for the global hospitality app, inHouse. By using Pepper, members can ask questions about restaurants in the network and make reservations without having to interact with a human. This saves the program and restaurants the overhead of having to pay a team of employees that would otherwise have to perform these tasks that simple for the AI to handle on its own. Pepper even landed inHouse a the spot in the Webby Awards for AI category, ultimately receiving an honorable mention just as it did in the Food & Drink category!
The 24/7 availability of an AI chatbot helps drive customer engagement by providing a system that’s “always on.” The clothing retailer H&M recently introduced an AI chatbot function in Kik messenger that reduces the need for customers to manually browse the retailer’s website or visit a physical location where particular items may or may not be stocked. North Face did something similar by providing a tool that questions customers how they’ll be using their apparel then makes recommendations based on the customer’s responses.
The cost savings are different for each individual company but ultimately, such systems inherently provide solid ROI by mostly eliminating a role that requires ongoing costs to support. By replacing humans with bots that can handle mundane tasks (and eventually, more complex tasks) retail businesses improve their bottom line.
For those interested in getting started, there are some great out-of-the-box solutions like Amazon Lex or the API from SnatchBot that can be used to build custom chatbots for customer service or really, just about any purpose. These free platforms give you a chance to add conversation capabilities to systems that can alleviate the need for human intervention in many situations.
We can implement AI in your project to save your business money
If you’re looking at breaking into eCommerce with an app, we suggest exploring the realm of AI to see how it can benefit your business. Get in touch with us at Blue Label Labs to learn more about our process and how AI can provide valuable solutions for your business.
Get the latest from the Blue Label Labs' blog in your inbox.
More in Development
SpaceX Software: Not so Different from Everyday Apps
At Blue Label Labs, we haven’t had the pleasure of sending any…
Moving Forward with Automation, Thanks to COVID-19
Many industries have already dipped their toe in the pool of automation…
Mobile Marketplace Apps: A Beginners Guide To Starting Your Build
The retail market was already moving to digital platforms in a big…