Image default
Guest Articles

Evaluating AI for Retail Pricing

Artificial intelligence (AI) and machine learning have left the hype cycle and are now being embraced by more than just the leading companies across industries. Not only that, having these robust data science capabilities is often considered critical for any business to establish and maintain a competitive edge.

In late 2018, we conducted a study with Forrester and found that 76% of surveyed leaders in the retail industry – a particularly competitive sector – believed that having AI boosts the organization’s bottom line. AI-based technologies have gained even more importance in the time since that research took place, which means the case for AI and machine learning in the retail context has only strengthened.

The motivating factor behind the rise of AI and machine learning technologies has been the explosion of data available to businesses. A common phrase in the retail industry explains that “retail is detail”. The source of this expression may not be known, but it gets straight to the heart of why AI holds such undeniable power for retail pricing. Think of AI solutions as this: those that make use of data to identify decisions are likely to produce the most beneficial outcomes.

AI Harnesses the Explosion of Retail Data to Get Pricing Right

In the case of pricing specifically, favorable outcomes include those that minimize cost and customer attrition, and reduce the risk of basket abandonment. Customer satisfaction is maximized, response rates are improved, and revenues grow, all thanks to data-informed pricing decisions. All this requires a deep understanding of the factors that determine customer purchasing behavior. Pricing dynamics ­– including promotional activity and inventory markdowns – increase retention and build customer loyalty.

Pricing is also a major contributor to margin leakage, something that businesses can’t afford to go undetected when there’s so much on the line, especially now in pandemic times. For all of these reasons, it makes sense why pricing optimization is regularly at the top of retailers’ list of challenges to tackle. And through AI, making optimized pricing decisions is more than possible.

AI and machine learning technologies supportgood pricing decisions by quickly and thoroughly finding the most important lessons the data holds. The massive datasets are the “details” that are so important to pay attention to for a thriving retail business. Data can propel a company from a state of analysis paralysis into the world of predictive and prescriptive pricing. If businesses have the ability to ingest and make sense of the data, they can extract key insights that directly inform decisions big and small.

Agile cloud-based technology, fueled by AI and machine learning, removes the need for costly internal infrastructure and resources previously required to support pricing and promotional practices, quickly leveling the playing field in a hypercompetitive landscape – or helping leaders solidify their position.

Three steps to Achieving Data-led Pricing in Retail

True AI-driven pricing optimization, like many new businesses processes, can only be enjoyed after completing AI baby steps in a “crawl, walk, run” evolutionary model:


In this beginning stage, retailers’ eyes are opened as good analytics (AI) reveal counter-intuitive insights. They may find that a large majority of promotional revenue comes from just a small fraction of promotions, or that the promotional lift they might see in traditional reporting comes unknowingly from product cannibalization. Retailers will have a clear-eyed view, maybe for the first time, of how effective their pricing and promotions really are. AI uncovers what worked, what didn’t and why. With this information, retailers can build solid pricing and promotional strategies – driving margin, growing basket, generating traffic, enhancing price image – and know which products can help achieve them. 


Retailers then graduate to walking. At this stage, they begin to implement rules-based price management and forecasting. Here, they fine-tune pricing best practices, determining good-better-best relationships, product families and promotional groups. Advanced data science allows retailers to see what will happen if they change a specific product price and what reaction it would likely trigger amongst a certain subset of shoppers. By doing this faster and more accurately than pricing experts could do with spreadsheets or other inefficient models, retailers can confidently test different scenarios and execute the best option.


When using AI at this step, pricing teams can see how demand and overall retail performance will be affected by any price going up or down and understand the scope of consumer price sensitivity and item elasticity on an ongoing basis. Because AI is an always-on solution, it constantly monitors market shifts and learns the dynamics of a retailer’s particular product assortment and shopper demographics. Here, the optimal everyday price is determined, along with ideal markdown pricing and promotional approach. The information from machine learning and optimization technology can also recommend targeted, prescriptive promotions, including which exact products to promote and which channels to promote them through.

With the ROI from each stage of implementation funding the next, and trustworthiness building with each profitable recommendation the solution provides, retailers can generate real momentum in their AI-enabled pricing fluency. The pricing and promotion space is one of great real-world consequence for retailers, which is why a considerable amount of thought and, you guessed it, detail, need to go into pricing decisions. AI and machine learning are technologies fit for the task.

Related posts

5 Reasons to Be Passionate about Coding

Asaf Darash

5 Ways Ecommerce Experiences Must Evolve to Succeed Among Economic Uncertainty

Jan Soerensen

Overcoming the Barriers of the Physical World with AI

Jonas Angleflod