Intelligent Tech Channels Issue 88 | Page 17

INDUSTRY VIEW
There are three types of AI that can help determine and execute effective pricing strategies: Generative AI, Machine Learning algorithms, and Cognitive AI models.
Well-trained Generative AI models have proven to be successful at predicting pricing trends, generating alternative pricing strategies, and simulating the impact of different promotional strategies on consumer behaviour. They have also impressed many retailers with their ability to generate and test alternative pricing strategies before someone commits the team to executing those strategies without evidence that they’ ll have any impact on buying decisions.
In one example, a global fashion retailer used Generative AI to create personalised discount strategies for Black Friday promotions. The result: a 12 % lift in revenue and 9 % improvement in profit margins compared to previous years.
AI analysed historical sales and past promotional performance, competitor pricing trends, customer purchasing behaviour by region, and also generated multiple pricing campaign simulations, example,
flash sales, tiered discounts, and personalised pricing and identified the optimal discount structure.
Similarly, some retailers are using Machine Learning algorithms to continuously refine pricing models based on historical sales data, past promotions, and real-time changes in consumer demand. One of the big benefits of Machine Learning pricing models is that they improve over time by continuously analysing large datasets to refine pricing decisions.
This has helped a high-end sneaker brand effectively manage dynamic pricing for fast-selling inventory, including a limited-edition collection that was launched despite high demand uncertainty. Traditional pricing would have used fixed markdown schedules, potentially leaving revenue on the table.
However, the Machine Learning-driven pricing model detected an initial surge in demand, predicted a sales plateau at current prices and recommended gradual markdowns instead of deep discounts. The result: real-time pricing optimisation led to a 15 % higher sellthrough at full price and a 10 % increase in total revenue.
Could a human deliver the same results on the same timeline without AI assistance? Probably not.
Finally, there is a growing need to use Cognitive AI, which considers external influences such as economic conditions, competitive pricing, local events, and shifting consumer sentiment to enhance pricing decisions. Integrating external variables such as economic shifts, local events, and social media trends without Cognitive AI is unfeasible, and these are highly influential factors on buying decisions and, therefore, pricing.
There was a national retailer selling winter coats which faced challenges in optimising markdowns due to regional weather variability. The retailer decided to use Cognitive AI to integrate real-time weather forecasts, identify areas expecting colder-than-average temperatures, and adjust pricing in regions where winter demand was likely to spike.
AI suggested the retailer delay markdowns in cold-weather regions, which it did. The result: the retailer extended fullprice sales by three weeks, improving gross margin by 7 %.
Retailers have a huge volume of highly valuable data at their fingertips waiting for the right AI model to turn into revenuedriving value. Visibility of data assets is elevated, and intelligent automation is made real with pricing that meets revenue and customer expectations. •
So how can teams work with AI to drive better pricing decisions and improve margins?
INTELLIGENT TECH CHANNELS 17