Retailers are focused more than ever on quickly adjusting to changing customer preferences and demand. Specialty’s Café and Bakery is a great example of a retailer that is using data to drive decisions related to product development and selection, inventories, staffing, and more to attract and keep customers.
For example, retailers rely on business intelligence (BI) tools to predict future demand for products around known factors such as special events or holidays. Purchasing decisions based on these predictions are usually made by buyers, and they use their experience and instincts to place orders. Introducing Artificial Intelligence (AI) capabilities into the BI software can remove these manual steps and human bias to uncover newer insights and improve business outcomes.
How often do retail analysts explore what they think are the key factors of an outcome and stop when they confirm the hypotheses, simply because it fits their theory? How many times might there be other more important factors affecting the outcome that have not been explored? When the data sets are large, with numerous attributes, users spend a lot of time slicing and dicing for newer insights or apply their original hypotheses to a subset of data. But are they missing opportunities for unexpected insights because of human bias or simply because the task of analyzing so much data is too daunting?
AI can remove the human bias and facilitates newer insights when applied to data sources for uncovering unknown factors that impact an outcome. For example, BI vendors are delivering smart analytic capabilities using AI or machine learning automation to speed up the analytic workflow and deliver insights that would be more difficult and time consuming for a human analyst. These technologies empower analysts with advanced analytic capabilities they didn’t have before, such as automatic model building against unfamiliar and non-traditional data sources, including socioeconomic data, customer-service inquiries, weather data and competitive data. This offers more accurate product demand predictions, all while reducing the dependency on highly skilled data scientists.
We spoke to Lead Data Warehouse Engineer Matt Riley at Specialty’s Café & Bakery, one of San Francisco’s largest and fastest growing companies, to share his vision of AI enhancing BI capabilities. With more than 50 restaurants in California, Washington and Illinois, Specialty’s has been dedicated to using Birst’s Networked BI platform to make data-driven decisions to improve customer experience and profitability.
1. Please speak to your mission statement and goals of the company.
Our CEO has communicated three key areas of focus this year. The first is customer experience. The second is developing four new products. The third is to create a five-year strategy to obtain funding from our parent company FEMSA to grow our business.
We rely on data and analytics to assist in all three areas. For example, we use Birst to bring to market new products based on the data collected from our e-commerce site, where we capture 280,000 ways customers can purchase our products through product modifications such as removing ingredients or adding ingredients. We blend that data with external sources, such as costing, to determine new, profitable products. Tuna was a popular add-on ingredient to salads, so we created a new salad on the menu that was based on tuna. Using data and analytics, we are able to create variety for our customers based on ingredients they like, which keeps them happy and loyal.
2. It’s an exciting time for data-driven retailers like Specialty’s Café and Bakery. You rely on data and BI and analytic tools to offer a better experience to customers. Please explain how BI is improving customer experience and business outcomes.
One example of how we have improved customer experience is being able to offer the best product selection, specifically for soups. In the past, we rotated approximately 30 different soups throughout the year based on a menu that had not changed in 10 years. Using analytics, we were able to narrow down the top six soups by location and time. We then blended weather data into our analysis to find correlations between soup selection and weather. With this information, we adjust each store menu to offer customers the best soup selection.
Having a consistent and best selection of soups keeps existing customers happy and brings in new customers since we are improving the likelihood that they will enjoy the soup selected. Another business outcome of this use case is an increase in sales and profitability. By offering a selection of products based on our analysis of what we know customers will like, by time and location, we have more sales and less waste.
3. BI and analytic vendors are increasingly investing in new “smart” capabilities using machine learning to automate analytic processes such as data preparation and discovering insights. What areas is most challenging for you and how do they negatively impact the business?
We would like to leverage more non-traditional data sources to find trends or patterns in the data to deliver a better experience to our customers. For example, we want to offer smarter product recommendations to our ecommerce customers. Today, we suggest products based on past purchase history. For example, if a customer purchased cookies in the past, that is a recommended add-on product. This can be a negative experience for a customer that may have changed diet and is purchasing a salad. A better recommendation may be a fruit salad or vitamin water for a customer purchasing a salad.
4. How do you think smart capabilities can assist with these challenges and find patterns in data?
The great thing about AI is that you don’t have to rely on intuition or gut to make decisions. You can’t argue with the data, but rather use the data to make better decisions. AI can tap into other data sources that we don’t normally take advantage of, such as Internet chatter or social media, to better understand consumer behaviors or habits in real time.
AI builds the model for us so that we can apply this to make the right suggested product on our ecommerce site. We could also apply AI to quickly adjust our product menu based on current events and trends. If the system or AI finds a pattern to uncover that multi-color quinoa is the latest popular ingredient in salads in San Francisco, we need to start the process of creating a new salad that includes multi-color quinoa and add this new product to our menu in stores in San Francisco. If AI tells us that people are getting bored of arugula in salads in San Francisco, we need to remove this from the menu. I envision AI helping us create the best products and services for our customers, without compromising profitability.
5. Lastly, can you tell us about Specialty’s partnership with Birst and the value it has brought to the business?
The amount of value is immeasurable. First and foremost, all users of the BI system use a single version of the truth. Now, our executive meetings are very productive since we don’t have to spend time reconciling numbers, but instead, make fact based business decisions.Birst dashboards offer complete and accurate visibility into store performance to help all levels of the organization drive sales and plan for growth. For example, regional managers have daily metrics for their stores and catering business to adjust staffing, promotions, inventory, and more. Executives have visibility to revenue contribution, across stores, to plan more accurately correct issues and plan for future expansion.
Mona Patel works in Birst’s Product Strategy team. With more than 20 years of experience building analytic solutions at The Department of Water and Power, Air Touch Communications, Oracle, MicroStrategy, EMC and IBM, Mona is now growing her career at Birst. Mona received her Bachelor of Science degree in Electrical Engineering from UCLA.