POSTED : March 7, 2019
BY : Concentrix Catalyst
As brick-and-mortar stores experience increasingly trying times, the need to better predict consumer behavior has taken on new meaning. The rise of artificial intelligence (AI), machine learning and predictive analytics, along with readily accessible business intelligence (BI) tools, present opportunities and challenges for the industry. The speed at which retailers adopt and effectively use these new technologies can mean the difference between a targeted seasonal sale and a going–out–of–business sale.
Only a handful of years ago, demand forecasting required an advanced degree in statistics and experience with computer modeling. With the proliferation of easy-to-use BI tools, the average employee now behaves more and more like a data scientist. Having more citizen data scientists—employees that incorporate BI into their everyday work—can be a boon for retail, but it also presents unique problems.
Imagine store clerks outfitted with shoppers’ purchasing habits and measurements, able to streamline the consumer experience by greeting them at the door with a selection of products tailored to their size and desire. Men’s Warehouse is one retailer doing just that with proprietary software that helps shoppers identify matching accessories. It’s made possible with machine learning algorithms that process data about relevant products across all of its stores.
Not only is your average retail associate now outfitted with a range of analytic tools to help predict consumer behavior, but that data is also being fed back to corporate. If you take that data and multiply it across hundreds of stores, the wealth of information available can be invaluable in predicting industry trends.
However, quantity of data doesn’t necessarily generate clarity — or impact. All of this big data is being manipulated and visualized by employees across departments, in marketing, sales, finance, etc. When the executive team gets together to make informed decisions about the company’s future, leaders are now faced with competing interpretations of the same data. Data that once was going to enable the company to better serve customers through personalized shopping experiences is now creating confusion.
With a wide range of BI tools on the market, some companies have five or more BI tools per function. When you combine more BI tools with more employees using them, you have a recipe for multiple versions of the truth. Many companies run the risk of making a key business decision based on the wrong version of the truth and wasting effort, time and money on bad ventures.
In order to achieve growth goals, retailers will need to streamline the number of BI tools and create a strategy roadmap for their use. Increased centralization of data, improved governance models and AI and machine learning allow companies to move past human bias and the limitations of first–generation predictive analytics.
If predictive analytics can help brick-and-mortar stores identify future trends, prescriptive analytics takes that data and pushes the customer in a particular direction. To compete with online retailers where this is already in play, the in-store shopping experience has to achieve a similar level of acuity and personalization in the customer experience.
The use cases for prescriptive analytics in physical retail run the gamut:
To acquire new customers in an industry where many retailers are competing in a death match with online marketplaces, brick-and-mortar stores will need to be at the cutting edge of technology. As younger generations that have always had access to on-demand online shopping experiences take over the bulk of discretionary spending, retail leaders can’t ignore the future if they want to establish lifelong customers.
To see how we helped REI gain an edge with greater visibility into consumer behavior and buying patterns, check out our case study.
Tags: adobe experience, Analytics, BI, Big Data, citizen data scientist, data scientist, deep analytics, Digital, Intelligence, Machine Learning, predictive analytics, prescriptive analytics, real-time data