Halo Merchandising can provide retailers the opportunity to double company net income. The Halo Effect is caused by consumer behavior. Grocery consumers buy full meals, not individual items. Fully understanding the Halo Effect and more importantly realizing its potential benefits requires sophisticated AI technology and automation. The Halo Effect should be understood and leveraged in all aspects of merchandise planning from promotion, pricing, assortment planning, and space planning to forecasting and ordering. Using the Halo Effect in all aspects of merchandise planning ensures that a retailer will meet the needs of its customers.
Why now? Margin pressures are challenging brick-and-mortar retailers like no other time in history. Though COVID-19 has provided a temporary respite for many grocery retailers, as we start to climb out of this pandemic it is certain that the competition for sales and market share will return to an even more intensely competitive environment than what it was pre-pandemic. Restaurants will reclaim some or most of their business back, customers will have less income and retailers will have shifted a good part of their sales to much less profitable on-line channels.
McKinsey estimates that 30-40 percent of merchants’ current head office activities are automatable just with today’s technology which grows in capability every day. Simply automating the current processes without changing the intelligence behind decision making will continue to leave brick-and-mortar retailers at great competitive risk from internet retailers like Amazon which use AI as a key input to merchandise planning and decision making. The current push to automation provides an opportunity to update traditional category management processes and activities which have mostly remained unchanged for the last 50 years.
Daisy Intelligence’s whitepaper explains the Halo Effect, how it can impact all aspects of merchandise planning and why automation is ultimately required to maximize its potential.
The ‘Halo Effect’ is Caused by Consumer Behavior
Retailers are very familiar with the fact that certain products have adjacencies or strong affinities that cause products to be purchased together. A few familiar examples: Customers who buy soda pop may also buy potato chips. A customer who buys paint will also buy paint brushes, rollers, drop sheets, masking tape, etc. We refer to this positive Halo Effect as associated sales. Potato chips are associated sales to soda pop when they occur in the same transaction together.
Cannibalization is another commonly understood Halo Effect. A customer purchasing Coca-Cola is less likely to purchase Pepsi. Cannibalization is caused by customer choice and typically occurs with products in the same category or related categories. Cannibalization is a negative Halo Effect.
A third commonly understood Halo Effect is pantry loading or pull forward. Customers may stock up on items that are on promotion and be out of the market for future weeks reducing future demand. Pantry loading is a negative Halo Effect.
Products that have a long-life cycle like condiments, which may take weeks for a customer to consume, or product categories like general merchandise will take a customer out of the market for weeks, months or even years. Product life cycle can have a negative impact on future demand.
Associated sales and cannibalization effects in grocery typically occur within a single shopping event or transaction. In some categories like general merchandise and hardware, these effects can occur over time. A customer renovating their home will not buy all the required products at once, so key items can drive a sequence of item purchases over several transactions that complete a consumers desired solution.
The source of Halo Effects is consumer behavior. Halo Effects are due to the fact that customers buy product solutions, not individual items. Some product solutions consist of many items like the example of a pasta dinner. Promoting ground beef will cause some consumers to buy pasta, tomato sauce, parmesan cheese, bread, wine and salad fixings. Buying a suit will typically cause some customers to purchase shirts, ties, socks and shoes to complete an outfit for a special event. Some products are part of small solutions. For example, bottled water. A customer does not require other products to complete the solution for bottled water, you simply drink the bottled water.
A typical retailer will qualitatively understand all of the Halo Effects described here, but it is beyond human ability to enumerate all of the Halo Effects for all of the thousands of products in a typical grocery, pharmacy, hardware, apparel or other retail store. To implement these Halo facts requires quantitative understanding. It is not sufficient to understand that promoting soda pop will also grow potato chip sales. To maximize the opportunity, we should understand:
- Which brands of soda pop will drive which brands of potato chips and by how much?
- How do these Halo relationships vary weekly over the course of the year?
- What other items does soda pop drive?
- Do recent promotions affect the current weeks Halo?
- What are the cannibalization effects caused by each brand of soda pop?
- What are the cannibalization effects caused by each brand of potato chips?
- Do the cannibalization effects vary weekly over the course of the year?
- What is the pantry loading effect cause by promoting each brand of soda pop?
- Does pantry loading vary weekly over the course of the year?
- How do all of the above Halo relationships vary regionally or by store?
Understanding each of the questions for a single item, let alone thousands of items is not possible without the use of sophisticated AI and automation technology. An army of data scientists with desktop or cloud-based AI tools could not enumerate the answers to all the questions above for all of a retailers’ products, let alone compare and optimize which Halo Effects to action each week in all of a retailers’ merchandise planning processes.
To maximize the Halo Effect requires understanding and actioning all the effects, not just a few adjacencies per category.