Was this email forwarded to you? Get Behavior Analytics​

In 2014, Tom Davenport gracefully met with me.

Professor Davenport was already well-known for his book “Competing on Analytics.” The challenge facing the tech companies and their clients was adapting from descriptive analytics to predictive analytics.

Translating that analytical shift into a product got me into the Stanford University Vision Project sponsored by Andrew Ng.

In 2025, we are experiencing the transition into perspective analytics.

In retail,

Most technological investment is made in the backend systems, supply chains, and the digital world. Yet, it is worthwhile exploring the shift in physical stores.

Today, let’s go beyond the cliché of the ‘store of the future’ into a free-thinking exercise about the impact of prescriptive analytics on retail, specifically on stores.

(Thanks to Paula Rosenblum, who started the conversation!)

🤔Descriptive, Predictive, Prescriptive Analytics

Let’s start with definitions.

Descriptive Analytics focuses on what happened, specifically analytics historical data. For example, by analyzing sales data, the reporting and query system will share which products were best sellers by season and geography.

Predictive Analytics uses historical data to forecast future trends. For example, the system will predict customer engagement with products in the store by analyzing the customer flow and point of sales data.

Prescriptive Analytics predicts outcomes and recommends specific actions to maximize results. For example, the system will offer an optimized planogram to maximize category sales.

Here are three scenarios where prescriptive analytics can play a role:

Dynamic Pricing

• Trigger: The system monitors real-time foot traffic, merchandise, and point-of-sale (POS) data. By 2 p.m., sales are 40% behind the daily quota.
• Alert: The system alerts the manager that foot traffic is less than planned. To reach the sales quota, the system recommends a pricing promotion (for example, buy two for the price of one).
• Response: the system monitors the sales results and suggests ending the pricing promotion once foot traffic picks up later.

Out of Stock

• Trigger: Sales are less than forecasted for a particular item. A camera solution identifies that the item is not on the shelf, and the inventory management system notes it is in the store.
• Alert: The system sends an alert to store associates' mobile devices that the item is out of stock on the sales floor and must be replenished from the back room.
• Response: The system monitors task compliance and sales.

Queue Management

• Trigger: The queue management system identified a group of shoppers arriving simultaneously at the checkout zone. The system recognizes a couple (shopping unit) and 3 individuals. That is higher than the checkout policy of ‘no more than one person waiting in line.”
• Alert: The system alerts the store manager and an associate on the floor to open another cashier station.
• Response: The system monitors the employees’ response time and how it impacts the queue flow, the number of people waiting, and their waiting time.

You probably noticed the *Rapid Response* trigger, alert, and compliance structure. It's designed for in-store growth loops.

😀Qualitative and Quantitative Research

The premise behind Behavior Analytics is to merge the depth of qualitative research with the breadth of quantitative data.

In other words, the Quali-Quant methodology aligns the Voice of the Customer with the scientific ability to replicate outcomes.

You can also leverage the direct and intentional information that defines zero-party research in physical stores.

👉 The Quali-Quant-Test Trifecta is powerful. 👈

While predictive analytics generates recommendations and actions based on quantitative data, the store offers an opportunity to talk to the shoppers directly.

Let’s take our three scenarios and search for behavioral nuances.

Dynamic Pricing

Predictive Analytics evaluates pricing and promotions based on historical sales data. It looks at what worked and what did not.

Yet,

Sales are outcomes.

Remember,

Sales = Traffic x Conversion x Basket

Say you sell apparel, and the conversion rate is 21%. The obvious step is understanding why almost 8 out of 10 shoppers left your store without a purchase.

The common problem in selling apparel is sizing. Your customers have various sizes, heights, and body contours.

Not all customers are the same.

Add trends, colors, and product placement, and you have a serious challenge aligning the customer’s desires with what she finds in the store.

In this case, pricing promotions can help alleviate the customer's dissatisfaction by NOT finding what she wants.

The associate can insert the information to get targeted pricing promotions and improve customer experience. At the same time, the system has details of why the purchase did not occur.

Even if a purchase didn’t happen during this visit, it is still a Win-Win.

Out of Stock

An out-of-stock event is a function of store operations.

Task Management systems focus on sending instructions and compliance rates. Yet, that’s an incomplete, sometimes even harmful, way of thinking about the event.

You want to understand what happened.

For example,

• An employee called in sick, and the store was short-staffed,
• The previous shift messed up the delivery process.
• The staff was busy serving customers, which took priority on doing the tasks.

Understanding why the out-of-stock scenario happened in the store and differentiating between corporate and local challenges is a win-win for the organization.

Queue Management

There are always tradeoffs when it comes to improving customer satisfaction during the checkout process. For example, you can invest in mobile payment systems, self-service kiosks, or open more cashier stations.

Yet,

Often, retailers miss behavioral nuances,

For example,

If you have been reading the newsletter for a while, you already know that I dislike the question, “Did you find everything you wanted?”

Don't ask the question if you don't give the associate a way to deal with the customer’s response. It backfires. 💣

Once you nailed the Quali-Quant-Test process, the challenge is managing it. That’s why you need Store Optimizers.

🤩Prescriptive Analytics for *Store Optimizers*

In essence, the AI revolution is about moving to predictive analytics solutions. In other words, we are moving from a world where tech is a tool to where it decides for us.

Practically, we are seeing a product market fit collapse.

The Product Market Fit theory depends on customer expectations, and because AI has accelerated the rate of market change, the threshold is steeper. ​

Here are the seven shifts per Reforge:

• "A Place For Me To Create" → "Do The Work For Me"
• "One Size, I Customize" → "Custom Made For Me"
• "I Expect To Wait" → "I Expect It Now"
• “I’ll do the busy work.” →“The busy work is done for me”
• “I’ll Pay Per Seat” →“I’ll Pay For Output”
• “The tool has no context” →“The tool can see what I’m doing.”
• "I'll Learn This Interface" → "The Interface Adapts To Me"

In other words,

You need to understand what it means to have an expert who can help everyone in your organization work more effectively.

Here’s an example from McKinsey:

Let’s look at our three scenarios:

Dynamic Pricing

Pricing is a value play. It depends on how consumers perceive your product's value (technically, Jobs-to-be-done).

Pricing promotions are about FOMO (Fear of Missing Out).

The role of a Store Optimizer is to know the differences between value and FOMO. Specifically, their job is to align pricing to the store scenarios.

For example,

Say a store associate discovers that a shopper likes a dress, but it doesn’t fit her properly, and the price is too high.

In that case, the store associate can suggest customer tailoring services, similar dresses, or discounts on an online order.

The Store Optimizer tests which scenario best balances the profit outcome and customer satisfaction.

Out of Stock

If you evaluate the out-of-stock event within Task Management, you face the challenges of Time Scarcity and Compliance Rate.

Time Scarcity relates to managing priorities. For example, Home Depot's policy says helping shoppers is the priority. Yet, if you don’t have a way to monitor what employees are doing, the reliance only on a Task Management System will backfire.

The Compliance Rate is in danger of becoming one of those KPIs everyone knows about and hardly understands. Nuances matter.

The Store Optimizer tests tasks' impact in the context of what happens in the local store and the corporate priorities.

Queue Management

The Peak End Rule is the most powerful consumer emotion retailers need to manage in the store, and good store managers know this well.

The Peak-End Rule refers to a psychological principle in which people judge an experience largely based on how they felt at its most intense point (the peak) and at its end rather than on the average of every moment of the experience.

The Store Optimizer is responsible for the deep dive into the nuances of Queue Management and the impact on customer satisfaction.

In short,

Prescriptive Analytics systems will improve decision-making in retail organizations, but, of course, the devil is in the details. 🤣

Fun!

P.S. I recently spoke to over 150 people to understand the best way forward for Store Optimizers. Below is the updated welcome survey that helps me create content such as this newsletter.

Thanks for filling out the survey. Click here. It took me 53 seconds.