February 1, 2021

How Data Analytics Empowers Retail Merchandisers to Thrive Through Disruption

No items found.
How Data Analytics Empowers Retail Merchandisers to Thrive Through Disruption

If 2020 has taught us anything, it’s to expect the unexpected. Wildfires, hurricanes, civil unrest, economic uncertainty, and a global pandemic disrupted businesses across industries — and retail was no exception.

Brick-and-mortar retail chains saw their foot traffic plummet, online stores buckled under the strain of increasing transaction volume, and both had to deal with supply chain disruptions and the fallout of the US Postal Service crisis.

But every challenge brings opportunity. As the Stoic maxim goes, “The obstacle is the way.”

Merchandisers live at that critical intersection between retailers and consumers, supply and demand, and inventory and marketing. Savvy retail merchandisers are in a unique position to help their organizations navigate through disruption, mitigate losses, and even increase profits.

The rapidly changing retail landscape demands that companies be able to pick up on rising trends and patterns and change tactics at a moment’s notice. To do this effectively, modern data analytics are the key.

Data opens doors (and can keep the doors open)

While many of the platforms available to retailers can generate massive amounts of data, most retailers lack the analytics tools and know-how to translate that data into actionable insights. More often than not, data isn’t used to drive day-to-day decision making, but relegated to executive-level requests and one-off projects.

For retailers to really reap the benefits of their data, they must make near-real-time analysis accessible to as many of their decision-makers as possible. Every item, every transaction, every touchpoint, and every customer has the potential to yield rich insights that can shape direction and strategy. From the way individual stores are configured to the brands and SKUs that are carried each season, retail merchandisers can use data to help their organizations boost their bottom lines and minimize losses — no matter what disruptions come their way.

Here are 3 adaptable plays retail merchandisers can execute to help their organizations weather any storm and even come out on top.

1. Product Affinity and Market Basket Analysis

What it is

Product Affinity and Market Basket Analysis are data analysis techniques that retailers use to uncover the relationships between items. By analyzing the contents of transactions, retailers can uncover combinations of items that occur together frequently and better understand correlations between them to develop better cross-sell strategies.

Why it’s important

Larger retailers have been conducting market basket and product affinity analysis for quite some time, but today’s fast-moving and unpredictable retail environment has only increased its importance as consumer tastes have grown even more fickle. Additionally, easier and cheaper data mining tools have opened up this kind of analysis to smaller and mid-size retailers, enabling them to better compete with big box stores. A few advantages of product affinity and market basket analysis include:

  • Targeted and personalized marketingBy understanding product relationships and purchase sequences, companies can identify and track customers who have bought similar products to deliver tailored messages and more effective marketing campaigns.
  • More effective cross-sellingRevealing the connections and correlations between product groups helps stores better improve their recommendation engine rules and algorithms to increase conversions. It also helps merchandisers understand which products to pair together physically or online.
  • Better merchandisingWhen consumer purchase patterns and product affinities can be identified and acted on, stores can create much more effective product promotions, shelf layouts, aisle displays, and more.

Why it’s hard to do now

  • Overwhelming volumes of siloed data.The explosion in the volume and variety of data retailers are able to collect is a two-edged sword. Purchase history, transaction-level data, and product information are usually spread across multiple, discrete systems with no clear or easy way to integrate them.
  • Compute power and scalability issues. Hundreds or thousands of products can easily generate billions of data points that take a staggering amount of computation power to analyze. Spreadsheets out near a few hundred thousand rows, requiring multiple spreadsheets and data extracts to keep up. This also results in only a subset of the data being analyzed at any one time, preventing merchandisers from getting the full picture. Capturing real-time PoS data, especially during peak seasons like Black Friday or Christmas, pushes systems to the brink.
  • The need for speed.Even if systems are integrated and compute power is available, complicated BI tools that require coding and technical expertise usually leave decision-makers at the mercy of BI and data teams. As a small group supporting the diverse needs of an entire organization, it can take days or even weeks for BI teams to give retailer merchandisers the insights they need, causing organizations to miss out on the impact of timely insights. In the world of retail, decisions need to be made by the hour, not the week.

How to do it better

  • Tap into a single source of truth. PoS data, online transactions, product information, clickstream, loyalty program, and omnichannel customer data can all be fed into your BI team’s cloud data platform, like Snowflake’s Data Cloud, providing your organization with a single source of truth. Multiple data formats can be stored and analyzed together, eliminating data silos.
  • Embrace the cloud. Spreadsheets were not built to handle the volume and velocity of big data being generated by today’s applications. In contrast, modern data platforms that were designed to leverage the unlimited scale and speed of the cloud allow companies to seamlessly store, manage, and analyze billions of live data points in real-time.
  • Get direct, real-time access to the data you need. There’s a new generation of analytics solutions that allow business teams like merchandisers and marketers to analyze and explore data at the same level as analysts — without having to know code. These tools take the form and function of tools merchandisers are already familiar with, like spreadsheets, to allow them to independently generate insights and take action without waiting on BI teams — all without sacrificing security, governance, or compliance.

Sample scenario:

A rare October snowstorm is forecasted to blanket the Northeast. While this would usually be a good opportunity for stores to advertise winter toys like sleds, a flu outbreak in the area has caused parents to restrict their children from going outside and interacting with each other. Instead, merchandisers at local stores dig into historical data to uncover the most common food pairings purchased during the winter season over the last five years. They shift their focus on creating displays with comforting treats that can be enjoyed at home, highlighting frequently co-occurring products like hot chocolate and marshmallows and apple cider and cinnamon sticks.

Dig deeper

1. EXAMINE A LARGER WINDOW OF TIME. While impactful patterns can be found at the weekly or monthly scale, be sure to zoom out further for seasonal trends or those that have slowly grown over years. For example, milk from non-dairy sources has been steadily rising in popularity over the last several years. Initially thought by many to be a passing fad, sales of oat milk have been up 289% year-over-year and have rapidly accelerated through the COVID-19 pandemic. Grocery store chains with the ability to analyze data over time can use this information to merchandise stores and advertise accordingly.

2. ANALYZE ABANDONED CARTS. Incomplete online checkouts and what’s inside these abandoned carts can reveal many areas for improvement, including product discoverability, recommendations, payment processes, website speed, and more. In fact, it’s estimated that $260 billion worth of lost orders are recoverable solely through a better checkout flow and design.

2. Product Mix and Placement

What they are and how they’re related

Product mix or product assortment is the total number of product lines a company offers to its customers. This includes the total number of products in a company’s product line a store chooses to carry as well as all of the variations of those products. Product placement refers to the strategic placement of products displayed in-store or online.

Why it’s important

Nailing the right product mix or assortment can have a big impact on sales. It’s critical to achieve a good mix of items across price points and preferences while not bombarding customers so they feel choice paralysis. Placement is another challenge, as retailers strive to maximize profit and capacity.

  • It shapes your brand image and the customers you attractThe wrong product mix or assortment can alter the image of your brand and the types of customers you want to reach. For example, suppose a store like Neiman Marcus carried only bargain brands or inexpensive items. In that case, they could easily lose their reputation for being a luxury retailer and cause their clientele to turn to competitors.
  • It makes it easy for customers to buyThe rise of online shopping has transformed consumer behavior. Customers come with clear search results, demands, and expectations about the products they’re looking for and will quickly move on if they feel they’re not being met. Ensuring your store, whether physical or online, is immediately surfacing and effectively displaying the products your customers are looking for is a constant challenge.
  • The paradox of choiceWhile it’s important to give your customers a range of choice and variety, too many choices overwhelm shoppers and ultimately reduces conversion. In one study, a display of 24 jams and jellies was put out on a store floor. While it attracted shoppers, only 3% converted. The next week, a display of only 6 jams and jellies was created. While that display drew less shoppers, 30% converted.

Why it’s hard to do now

  • Top-selling products aren’t always the most valuable.When determining which products should occupy the most shelf space, more expensive items aren’t always the biggest moneymakers. Margins, rate of sale, returns, and other factors can all affect the amount of revenue each item can drive. Some key questions to ask are: Which products drive the most loyalty or repeat purchases over time? How much does it cost to acquire the customers that purchase these products? What are our margins on this product line?
  • The dashboards your BI team provides won’t cut it. With dozens or even hundreds of products in a product line and multiple variations of each item, analyzing data in aggregate can’t provide the insights you’ll need to determine which product assortment will drive the most revenue for your store. You need to get granular and analyze things like individual SKUs, transactions, and customer ID in real-time. Unfortunately, the dashboards your BI team provides don’t let you drill into that level of detail without asking for help.
  • Supply chains can be interrupted. Disruptions like bad weather, material shortages, COVID-19, regional conflicts, etc., can affect supply chains and restrict availability. Retailers may need to draw up multiple contingencies to address any disruptions to supply.

How to do it better

  • Get a clear picture of your best customers. The total amount of monetary value a business expects to earn from a customer over the entirety of the relationship is life-time value (LTV). Target segment analysis can help you determine which groups of customers generate the highest LTV and analyze their transactions can reveal which products they have an affinity for. Use these insights to guide your product assortment and placement strategies to drive maximum revenue.
    For step-by-step instructions on conducting Target Segment Analysis, see this Marketing Play – Identify Your Highest Value Shoppers for Maximum Profitability
  • Get granular. Nailing the perfect product mix requires going beyond the overarching metrics in your BI dashboard. Retail merchandisers need to dig into overall sales by product line to examine performance by particular SKUs, splice that data further by region, and so on. Of course, this requires a cloud-native tool with unlimited scale and speed.
  • Become a data explorer. Digging into data and coming up with new and creative solutions to increase revenue through product mix and placement requires the ability to do rapid, iterative analysis. Retail marketers and merchandisers must be armed with analytics tools that empower them to independently explore data and get answers to their questions in real-time. Questions like, does grouping the sprinkles with the chocolate syrup, chopped peanuts, and canned cherries have a greater impact on sales than grouping it with the cake mixes and frosting supplies?

Target Your Highest Value Prospects for Maximum ROI

Step-by-step instructions on conducting Target Segment Analysis


Sample scenario:

An influencer on Instagram raves about her favorite ice cream brand to her 4 million followers. Based on her endorsement, demand for that ice cream brand skyrockets across the country. To best capitalize on this trend, local merchandisers analyze the top-selling flavors in the product line by region to determine the best flavors to promote for local palettes. As a result, “Praline Pecan” becomes a big hit in the southern states while “New York Cheesecake” sells out in the northeast.

Dig deeper

1. INCORPORATE SOCIAL/ENTERTAINMENT TRENDS. Box office data, Niesen ratings, and social trends can provide opportunities for revenue driving product placement. Sales of chess sets exploded after “The Queen’s Gambit” topped the Netflix chart. Upcoming sports events can be a source of inspiration as well. Are the Giants in the World Series this year? Orange and black napkins are sure to be hot sellers for baseball fans throwing viewing parties.

2. CONDUCT SEASON-SPECIFIC ANALYSES. Seasonal shoppers tend to spend more during particular times of year because they focus on products related to a specific event, holiday, or season. Zoom in on their preferences and purchase behaviors to guide your seasonal inventory, ad targeting, and displays. Some seasonal assortment and placement concepts include:

  • Seasonal recreation products (i.e. pool toys in the summertime)
  • Holiday items (Christmas trees, turkey platters, etc.)
  • Items that match seasonal colors (i.e. reds, oranges and browns in fall, white and blues in the winter, bold colors in spring, yellows, oranges, and greens in the summer)

3. Recommendation Optimization

What it is

Recommendation engines are designed to present products and services that customers would likely be interested in based on past purchases, interests, behavior, season or other relevant factors. This personalization is an effective way to increase conversion.

Why it’s important

Digital product recommendations are designed to better meet the needs and desires of customers, as well as boost sales. The best time to get a customer to convert is when they are already ready and willing to spend money. A well-tuned recommendation engine can generate huge profits. It’s estimated that Amazon.com’s recommendation engine is responsible for 35% of its revenue!

  • They increase sales and basket sizesRecommendations capitalize on prime points of influence during the customer journey. The right recommendation to the right shopper at the right moment can make the difference between an abandoned cart and a multi-item sale.
  • They drive retention and repeat purchasesRecommendation engines are a form of personalization, and personalization improves the customer experience. A positive experience brings customers back time and again so that even frequent customers who purchase lower-ticket items on average actually generate greater value than one-time big spenders over time.
  • They grow brand loyalty and evangelismCustomers trust retailers that consistently provide them with what they need when they need it. They are more likely to choose your brand over competitors, write positive reviews, or tell their friends about them.

Why it’s hard to do now

  • The customer journey is a labyrinth of tons of touchpoints.It’s hard to know key points of influence, opportunity, and risk when the customer journey has ballooned to an average of over 60 different touchpoints. With hundreds, thousands, or even millions of customers across all stages of the customer journey, identifying the right product to display at the right time becomes difficult to do at scale across millions of transactions.
  • Every customer is unique, and so is their data. Millions of customer data points are streaming in across PoS systems, loyalty programs, online payment processors, and dozens of other platforms. Consolidating this mish-mash of information in a way that gives marketers and merchandisers a meaningful picture of each customer and fuels relevant recommendations at scale is extremely difficult.
  • There are influencing factors outside of your control. Weather data, holidays, cultural and religious considerations, and other external data factors can play a large role in what customers want to see, experience, and purchase, But it’s difficult to integrate into recommendation engines effectively. And if the implementation is done poorly, the errant recommendations can turn off, offend, or distract the visitor.

How to do it better

  • Build a customer 360. Having a clear view into the customer journey and understanding every touchpoint across the funnel allows you to determine where each buyer is on their path to conversion. This insight helps you develop rules for your recommendation engine that improve customer interactions by providing the best possible recommendation for their current funnel stage. New users? You have minimal data about them. Display generic recommendations (ex. Most popular, top-selling, best rated, etc). Loyal shopper with a long purchase history? Personalized recommendations work best, followed by things like recently viewed or recently purchased.
  • Create customer profiles. Once you have the single source of truth we mentioned in the first play, you can begin to build out profiles with a definitive set of demographics, behaviors, and characteristics that allow you to deliver more personalized recommendations at scale. Look at folks with similar ages, zip codes, behaviors, etc., and recommend their purchases to one another. The longer your relationship with the customer, the most personalized and relevant your recommendations will be.
  • Integrate third-party datasets for even more impactful insights. Fully leveraging all of your data is a great first step, but leading retailers augment their data with third-party datasets for even more powerful data insights. External data can be used to build product association rules based on external factors that are having a direct impact on the individual shopper. Is a thunderstorm forecasted for tomorrow in Tulsa? Display umbrellas on the homepage for visitors coming from Oklahoma IPs. Snowflake’s Data Marketplace is one resource for live, ready-to-query datasets.

Sample scenario:

After analyzing historical transactions against customer 360 data, retail merchandisers discover that both men and women across several age brackets seem to purchase health and fitness equipment in January. They determine that this is a seasonal trend based on the popularity of New Year’s resolutions. They update their recommendation engine to suggest personal fitness products to all customer segments during the month of January, even if those segments had previously not expressed an interest in fitness before.

Dig deeper

1. SQUEEZE THE LEMONS. Just because an item sells well doesn’t mean it’s helping you turn a profit. Analyze the top SKUs that are returned and remove these from your recommendation engine. Or, heavily discount them and recommend them to customers searching for bargains. On the same note, avoid recommending sale or clearance items to your highest value customers. Instead, recommend your highest margin products to them.

2. AUTOMATION IS EASY, BUT HUMANS ARE BETTER. Recommendation engines are largely automated, which enables them to function at scale. But humans can pick up on trends to better tailor recommendations to customers. Humans can also quickly adapt and incorporate complex or changing business needs to help shape their recommendations. Choose a tool that empowers your merchandising team to explore and analyze data themselves without any help from the data team.

Navigate Through Disruption With Data

In today’s ever-changing retail environment, retail merchandisers can use data analytics to improve their store’s bottom line, no matter what disasters may strike. By gaining a deeper understanding of customer shopping habits, preferences, and optimizing their physical and digital storefronts based on these insights, they can increase sales, improve customer satisfaction, and maximize profitability.

Taking a data-driven approach across the organization requires a culture shift. Teams from across the organization, including marketing, merchandising, operations and finance, must work together to collect, analyze and act on the data. Fortunately, the right tools can enable real-time collaboration without adding complexity.

Sigma is a cloud analytics and business intelligence platform that’s empowering online and physical retailers of all sizes to harness the power of data to maximize their sales:

  • Faster decision making Visual information is much easier to process than written information. By using a chart or graph to summarize complex data, the audience absorbs it quickly, allowing business leaders across the enterprise to evaluate and interpret the data.
  • Identification of areas for improvement With the help of data visualization, organizations can see where performance is high, as well as where there’s room for improvement. For example, if your marketing team knows that for every X number of campaign emails, Y number of website visits will result, creating a visual report based on clicks per email, and progress to traffic goal is a visual motivator to meet the traffic quota.

Let’s Sigma together! Schedule a demo today.

We are Sigma.

Sigma is a cloud analytics platform that uses a familiar spreadsheet interface to give business users instant access to explore and get insights from their cloud data warehouse. It requires no code or special training to explore billions or rows, augment with new data, or perform “what if” analysis on all data in realtime.