How Analytics Helps Price Optimization Decisions
Money paper airplane in sky
Amazon's quick price cut for Whole Foods, and investor concerns about Blue Apron, together place a spotlight on price optimization in retail and eCommerce. I am going to try to explain some analytics reporting can aid good decisions, and support informed discussion, on product planning and associated metrics for online retail.
High sales have traditionally attracted attention in business news. But savvy business leaders are starting to recognize that while dollar signs fit a headline, optimizing prices is the best long-term differentiator. Widespread business intelligence is making analytics essential to the decisions behind price optimization.
Probably the sharpest example comes from eCommerce grocers, particularly the plight of Blue Apron, a meal-prep delivery service that recently issued an IPO. The public debut opened the startup to financial scrutiny of its cost structure. Analysts soon critiqued the high customer acquisition cost of BlueApron – $400 per customer according to Business Insider. Other similar online service-prep providers, looking to ride the eCommerce tide started by Amazon, have seen similar challenges in achieving good profit margins.
Service providers are realizing that competitive advantage with an offering must be paired with advanced techniques that reveal a better real-time understanding of consumer behavior. Personalized price optimization has become essential to a company's outlook as a result.
Price optimization has been gaining interest over the past few years, especially among service grocery retailers. A typical grocery store relies on volume sales of low-margin consumer goods. Personalized pricing – optimized pricing based on customer behavior and preferences – created a breakthrough in managing revenue rather than an concentrating just on sales volume. The results have gained a higher profile as more retailers have embraced the strategy. In 2014, the Star Tribune reported how the retailer Target had increased same-store sales because of its price-matching policy on food and health products.
An MBA student with a good SWOT can easily suggest price optimization as an essential element in determining profit. Analytics, like software, is everywhere, though optimization capabilities were different at its inception.
When web analytics first burst onto the business scene, many of the features and metrics related exclusively to paid advertising or ecommerce concerns. Experts conducted analysis by inferring customer intention from a given website element – such as assumptions based on shopping cart abandonment.
The consumer adoption of retailing via smartphone has altered that inference analysis. Customers once just shopped at home from a desktop or a laptop. But when tablets and smartphone were carried as personal items, customers discovered they were able to compare products and services while in store or on the go.
For analysts the new behavior extended the context of the reported metrics to a more detailed persona and deeper interpretation of what customers wanted. BOPIS (Buy Online Pay In Store) and BORIS (Buy Online Return In Store), in which customers shop at home and then pick up their item (or return it), has blended offline and online activity. This means analysis can more accurately understand the conditions in which online activity occurs.
So how can analytics help to make a business have better profits from future quarters? Analytics can provide the basic platform for gathering the metrics associated with retail activity and highlight where such activity holds opportunities for improvement.
Online/offline activity like BOPIS and BORIS has consequently led business leaders to think about their spend rate more critically. Customer transactions are occurring faster, with digital options such as paying online at a portal, and mobile payment readers from Square, Paypal, and other providers. The faster rate means businesses must inspect how transactions impact their inventory, operations, and ultimately cash flow with a greater frequency than traditionally believed.
Amazon's surprise decision to adjust prices for Whole Foods produce is a great example of the importance of speedy decisioning applied to the right inventory. The initiative has also driven competitors to look at their business intelligence to learn how to counter the move.
There are a few business intelligence frameworks that are good starting points for comparing analytics metrics to building models for price optimization value.
- What-if scenarios for cost
- Production yields
- Sales by customer and product
- Product margins
- Analysis of purchase decisions
Standard analytics solutions can assist with some aspects of these frameworks. The metrics can provide past data for predictive trends, encourage discussion about underlying assumptions supporting the trends, and enhance SWOT discussions based on digital media.
Advanced analytics can address the gaps. They allow the combination of data sources, leading to a comprehensive customer view. One useful feature that has been adopted in major analytics platforms is a User ID tag. The feature assigns a generic number to each visitor who arrives at a website page from one device to another. The purpose is to allow cross-device segmentation by behavior, while avoiding identifying an individual and breaching visitor privacy preference. Highlighting the behavior of a derived segment can give ideas about linking online behavior to pricing optimization opportunities.
Regression models are another way to learn how costs correlate with retail events. In R programming for example, a simple time series of sales can be plotted, then further analyzed with a decomposition analysis. The analysis is meant to separate and model the data into three graphs. One graph shows the general trend of the data. A second displays the spikes, providing the size and frequency. The third displays data that is considered noise. This analysis conveys how data spikes occur. If the data represents sales or activity, managers can envision how to better manage inventory costs that are a response to the spikes, and give indicators as to when price optimization can be used.
Getting the right price was once about just knowing how to value your product or service. Today it is part of the analytics narrative: A business that manages its assets best with business intelligence will outlast its competition. So far Amazon has brought that narrative to life. Being savvy with price optimization is one way new entrants can have the benefits of that same narrative.