This article is part of CMO.com’s November series about commerce and consumerism. Click here for more.
Amid a rash of high-profile brick-and-mortar store closures, it is registering with leaders in retail that they must refocus their sales and marketing efforts in order to survive.
The opportunity at hand: to transform the customer experience by embracing a new breed of technology—specifically, artificial intelligence (AI).
AI is no longer a futuristic technology that promises to transform businesses. It's a here-and-now operational tool that can help retailers manage costs, make their brands more relevant to consumer needs, and create more personalized experiences.
But AI is not like a software package that retailers can buy off the shelf, plug in, and run. To achieve the benefits of AI, retailers need to reorient around their data. One key to creating a data-driven foundation is to engage with data scientists. These are experts who possess a deep understanding of designing algorithms that support the needs of a business.
Non-retailers realized the value of data scientists early on, and many, such as Apple, Facebook, and Google, are enjoying a clear competitive advantage as a result. Even Spotify now relies on data scientists, which has helped the tumultuous music industry achieve its first period of growth since the 1990s.
Driving Operational Alpha
From a retail operations perspective, AI can help reduce costs and increase efficiencies. For example, by providing intelligent customer services through bots, or personalizing a shopping experience with more refined inventory recommendations based on shopper-provided preferences, AI can cut labor costs. It also can boost brand relevance by predicting customer-buying trends based on purchasing patterns and social media chatter.
Case in point: A leading retailer used AI to curate product images for its e-commerce website. Previously, people manually reviewed, tagged, and uploaded individual images to a content management system. The process was time-consuming and fraught with human error. By using AI, the company applied computer vision and machine learning to more accurately categorize product images in an automated fashion, resulting in a faster processing time. People are still involved in image curation, but only to make judgment calls, such as when the technology is unsure how to tag an item.
As a result, the retailer has cut costs for image curation online and has increased productivity, as employees are able to focus on a higher value-added activity.
Audience Insight And Engagement
As businesses achieve more operational maturity with AI, they can then apply the technology to gain stronger insights into audience preferences and behavior across multiple channels. Why? Because AI can analyze far-flung and seemingly disconnected sets of data more deeply and quickly than machines that do not teach themselves to learn.
For example, SapientRazorfish used applied machine-learning techniques to dynamically predict user behavior on a leading retailer’s website. Previously, the same retailer targeted predefined user segments with products and marketing based on a set of predefined rules. By employing the latest techniques in applied machine learning, the team accurately predicted the most likely department or product the visitor was looking for at that very moment. The ability to dynamically predict which products people actually need demonstrates the power AI brings to retailers today.
Businesses can also build on AI to improve customer engagement. Staples, for example, offers its business customers an Easy Button that makes it possible to quickly restock orders for office products. Pressing the Easy Button activates a voice interface that asks them what they need to reorder. By replying “Post-it Notes” or “blue pens,” they can restock on essential products—an elegant user interface akin to the Amazon Dash button with voice.
How can retailers get started with AI? Here are some initial steps to implementing AI as a competitive asset.
- Define your business problems: Achieve consensus around your most essential challenges and decide which ones can benefit from AI. Prioritize one problem first with a manageable scope to pilot, and then build consensus with key executives in your organization to champion a way forward.
- Define the data you need to solve a problem: This means taking stock of the business processes that can deliver immediate benefit if they were improved and defining the kind of data that you need to improve the process.
- Build a practice for acquiring and harvesting data: This is one of the ways data scientists can play a crucial role.
- Design a prototype for how you're going to collect and apply data to improve a process: Your prototype should validate whether you can collect and use data successfully. That means doing things such as running live data tests with models and collecting and analyzing data on the performance of the models you have constructed.
- Go live with an actual application of a prototype that will manage the data required to solve the business problem you set out to improve with AI.
A retail business with the right leadership and mindset can flourish with AI, but it must first build a culture of data-driven decision making. In today’s competitive landscape, it is essential that retail leaders consider the real impact that implementing AI solutions could have on their companies. Besides cutting costs, AI can provide invaluable audience insight, drive customer engagement, and accelerate sales.
No other technology currently holds as much promise to save the retail industry as AI.