Algorithm Price Wars: risk and opportunities for consumers

Technology and data science have fundamentally changed the way retailers do business, particularly when it comes to pricing.

Competitive pricing while achieving business goals is a core challenge for retailers. However, thanks to data science and other advanced technologies, price optimization has become easier.

AI-driven algorithms have made it possible to adjust pricing on a minute-to-minute basis within a retailer’s online store or at any one of their physical locations. In addition, algorithms are a pivotal cog in helping retailers both monitor competitor pricing movements and autonomously adjust in response. And this ongoing “competition” between retailer pricing algorithms has resulted in delivering better prices and value to consumers.

Or has it?

Given the ongoing transparency issues associated with AI algorithms and many other types of data science, there is growing concern that instead of lowering prices, competition between algorithms is actually driving prices higher or causing them to stagnate. For example, according to a recent study by Harvard Business School, entitled Competition in Pricing Algorithms, many smaller retailers may opt against reducing prices out of concern that companies with more sophisticated pricing technology – such as Amazon – will undercut them within minutes. This means retailers are commonly leaving prices flat instead of making them as competitive or budget-friendly as they could be, resulting in customer dissatisfaction and distrust.

Is there anything that can be done to fix this? Yes. Here are three actions retailers and their technology teams can do to restore customer confidence.


Focus on Transparency

AI may be good at delivering quick results, but how were these results derived? With black box AI, you will likely never find out. Making the best pricing decisions means you need to know exactly how each price was calculated.

This is why transparency is so key. Yes, AI may have delivered a lower price. But what implications does this have on the brand? Were there any biases in a given pricing calculation? Does this price mesh with the desired KPIs your company has outlined? These are all questions that are incredibly hard to answer with a black box AI, which is why transparency in pricing decisions can be so important to consumer trust.


Combining the Black Box with Actual Knowledge

Black box algorithms can get pricing wrong because they don’t necessarily allow for nuance. Therefore, making price changes primarily based on the repricing activity of competitors can deliver results found in the HBS study in need-based products, such as OTC allergy medicine where price is less of a competitive factor, while leading to a race to the bottom for want-based products, such as electronics where price is the main differentiator. In addition, if left unchecked, blackbox AI can create price gouging situations when demand for a “need” product — such as hand sanitizer — is at its highest. This can significantly damage customer loyalty as prices for these necessary products may inflate to levels 10 times — or more — higher than they usually cost — and retailers may not even know that the price gouging is happening because of a lack of oversight.

A better way is a hybrid approach that combines pricing algorithms with institutional knowledge so that need-based products can be differentiated from want-based products, as well as complementary and substitute products. It can also provide the end user of the AI system with more control, allowing he/she to define rules and parameters for products that fall in each category, including pricing consistency within your product catalog using more of a leader-follower approach to ensure you do not confuse your customers.  


Prioritizing Collaboration

Leaving a cordoned-off AI black box in control of pricing is like throwing a bunch of ingredients in an Instant Pot and then having the dish completed in a short period of time. It’s a “set it and forget it” way of pricing that leaves too many people in the dark across an organization. However, using a platform that offers one version of the truth for various departments – including, but not limited to, merchandising, marketing and strategy, and analytics – allows the entire organization to contribute to the formulation of competitive and profitable pricing strategies. When everyone participates and has visibility, it is easier to align strategies with the overall goals of the company instead of focusing only on one department’s goals. In addition, the combined brainpower of people in various departments can help other colleagues, as well as the hybrid solution, recognize potential roadblocks that a black box alone can’t.

Related resources

eBook – Private label brands in pricing: optimize now, build up over time
eBook – Price-indexes and marketing insights
eBook – Dynamic Pricing
eBook – The paradoxes of price image
eBook – Assess the maturity of your pricing
eBook – Taking advantage of Big Data to regain margins
eBook – 3 lessons learned on price image


Questions ? Want to know more about our solutions?

Follow us: linkedin twitter