Rethinking Pricing and Promotions

Eversight is a Business Reporter client.

AI-powered experimentation unlocks faster and more accurate market responses than traditional models that rely solely on history.

The consumer goods retail industry is facing a fundamental dilemma: Standard market models for pricing and promotions have typically relied on past models, but no historical data exists that can guide today’s quickly evolving strategy. The industry has long leveraged historical data to extrapolate what might work at a given moment in time, and while there may be value in that approach, what worked yesterday may not fit today’s context.

If the past two years have taught us anything, it’s that “now” is always changing. The dynamics at play in today’s retail environment reveal why the pricing techniques and algorithms the industry has traditionally deployed are no longer enough. 

Inflationary pressures have become a fact of life. Manufacturers are tasked with raising prices across their portfolios to previously unseen levels; retailers are facing an unprecedented frequency of pricing actions; and the lag between when manufacturers increase prices and when retailers are able to reflect them is contributing to elasticities that look different than observed history. As once-in-a-generation inflation translates directly to shelves, historical data can’t inform a successful strategy for never-before-seen prices or identify what works at new, higher thresholds. 

There is no longer “online” or “offline.” Digital connectivity has permeated physical retail, fulfilment options are exploding and there is one omnichannel funnel, and a supercomputer in every consumer’s pocket. Smartphones and heightened price visibility are rapidly changing the way consumers react to price changes and promotional offers. 

Shoppers aren’t linear or rational. Models for pricing and promotions have long been guided by the assumption that shoppers are consistently rational beings, but they’re far more likely to act on impulse. Their habits, preferences and priorities—not to mention their expectations—look different now than they did two years ago, and they’ll look different again in six months. Staying relevant requires more than repeating or modeling the past. 

In short, traditional, history-based methods for pricing and promotions just don’t make sense anymore. 

Testing is required in today’s world

At Eversight, we’ve often extolled the virtues of A/B testing. A quick look at other industries sees the concept being used to optimize websites, ad-tech algorithms and streaming recommendations. The underlying principles represent the basics of the experimentation-based approach that we see as critical to successful pricing and promotions in consumer goods retail today. 

Experimentation in this context means to continuously and automatically test prices and promotions directly with real shoppers and then measure how shoppers engage across the different variations. Experimentation draws information and insights from the revealed behavior of today’s shoppers—not from what they’ve done before or what they’ve said they’ll do—making it easier to accurately identify the prices and promotions that resonate best in the current environment while still delivering against business objectives. 

By letting shoppers choose what works best and when, and analyzing and acting upon that data, an entirely different definition of what’s possible emerges. This dynamic approach to testing creates clean, current datasets based on shoppers’ real perceptions, resulting in more robust data and ultimately better pricing and promotion decisions. 

Experimentation unlocks the ability to accurately align business strategies with how people actually make choices, producing prices and promotions that resonate with today’s shoppers and drastically boost performance. Experimentation-based technologies now exist, purpose-built for the consumer goods retail industry, that allow continuous discovery of new and better-performing prices and promotions. 

AI and machine learning unlock transformational growth

Retailers have been utilizing test-and-learn forever, but before AI and machine learning, there was no way to run millions of tests simultaneously. No human could design a complete array of tests that run all the time and are perpetually refined, so analysis relied on the natural experiments of the past. But yesterday’s A/B testing is simply too slow and unstructured, and it relies on a human to decide what to test. 

AI and machine learning are required to support and automate experimentation as it’s too complicated and difficult to keep up with today’s algorithms, bots and competitors that change prices repeatedly—and shoppers that can see it all. No one is better suited to make some decisions than the people who are in the business day in and day out, but the right answer isn’t always having an algorithm for everything—and AI-powered experimentation produces infinitely better data to analyze and inform decisions. 

The best outcomes are achieved when humans and machines work together, compounding their respective strengths. AI and machine learning, when applied through experimentation, unlock revenue growth by automating complex decisions, uncovering the prices and promotions that work best today and providing the necessary flexibility for businesses and strategies to adapt as dynamic markets shift. 

The only way to predict the future is to build it

Given the disruptions of the past two years, ongoing inflation and the shift to omnichannel, a forward-looking, experimentation-based approach to optimizing prices and promotions makes infinitely more sense than historical methods. We’re seeing a profound change not just in dynamic price movement, but also in promotions—and, ultimately, in loyalty and personalization as the industry evolves. 

So, how do you keep up with constant change in today’s complex world? One thing is for sure: The answer isn’t periodic reforecasting using two-year-old historical data. To secure the future, AI-powered experimentation is needed. 

This age of transformation calls for a relentless focus on improving shopper experiences and business performance, and that means breaking free of the limitations and data of the past. AI-driven experimentation at scale enables agile and proactive pricing and promotion decisions. Through this methodology and the technology that supports it, prices and promotions can be smarter and reflect shoppers’ habits, preferences and priorities today. 

Learn more about the power of experimentation at eversightlabs.com.

— Jamie Rapperport, Cofounder and CEO, Eversight

Image: Courtesy of Eversight