Can AI be used in Pricing? | eCommerce Matters Ep. 023

Listen Now

In this week’s episode, Philip and Rob look to define what actually is Artificial Intelligence (do you agree with their definition?), and how it can be best used to support pricing decisions. Spoiler alert - you don't have to jump straight into the deep end, with an all-singing, all-dancing, self-learning algorithm to get value from AI. Even basic applications can be very effective. Listen, to find out more!

🎧 Listen now on all major podcast platforms.

🎧 Listen now on Apple.

🎧 Listen now on Spotify.

Watch Now

Full Episode

What actually is Artificial Intelligence?

Does a rules-based system count as Artificial Intelligence?

Do you agree or disagree with us?!

Does Artificial Intelligence (AI) need to incorporate the ability for the machine to learn, in order to be AI?

Fun Video

For this week’s podcast props, Philip picked a car (well a tractor to be precise!), and Rob, brought a favourite childhood book to introduce today's podcast title.

Podcast Summary

The Limitations of Predictive Models

  • We acknowledged that using price elasticity calculations and predictive models in pricing decisions can be challenging.

  • Predictive models often struggle with non-linear market dynamics and fail to anticipate unforeseen events accurately.

  • Discontinuities and market disruptions can lead to unpredictable outcomes that traditional predictive models may not capture.

The Benefits of Rule-Based Approaches

  • We emphasised the advantages of rule-based pricing approaches, which leverage historical sales data and predefined rules to make pricing decisions.

  • By incrementally adjusting prices based on established rules, businesses can optimise their pricing strategies.

  • Rule-based approaches allow for more control and understanding of pricing decisions, especially in complex and evolving markets.

Challenges in Predicting Disasters and Chaotic Events

  • We discussed the difficulty of accurately predicting and preparing for disasters, crises, or other chaotic events.

  • Historical data from previous crises may not be sufficient to predict future events, as each disaster tends to have unique characteristics.

  • Algorithms trained on specific historical events may fail to account for new phases, changes, or unprecedented situations.

The Significance of Explainability in Pricing

  • We highlighted the importance of explainability in pricing decisions.

  • Having transparent algorithms that can provide insights into the causal links between pricing decisions and market dynamics fosters trust and enables better decision-making.

  • The ability to understand why certain pricing recommendations are made is crucial for businesses to validate and fine-tune their pricing strategies.

Suitability of AI in Pricing Decisions

  • We examined the suitability of AI in pricing decisions based on the scale and resources of different businesses.

  • While AI can be a powerful tool for automation and optimisation, smaller businesses may have limited ROI in investing heavily in AI techniques.

  • The level of sophistication required in pricing decisions depends on factors such as data availability, market responsiveness, and business objectives.

Return on Investment (ROI) and Effort Considerations

  • We stressed the need to evaluate the ROI and effort involved in implementing AI in pricing decisions.

  • Businesses should weigh the potential benefits of AI against other priorities, such as advertising, stock selection, or infrastructure improvements.

  • Depending on the business's level of sophistication and the market dynamics, the investment in AI may vary.

The Importance of Human Expertise

  • We underscored the value of human expertise in pricing decisions and the role it plays alongside AI.

  • Human decision-making, domain knowledge, and market research are essential components for effective pricing strategies.

  • AI should complement and enhance human decision-making rather than replace it entirely.

Stepwise Implementation Approach

  • We suggested a stepwise approach to implementing AI in pricing strategies.

  • Businesses can start with basic automation, such as competitive repricing or tracking, and gradually move towards more advanced techniques as their needs and capabilities evolve.

  • The stepwise approach allows businesses to assess the benefits and impact of AI at each stage and make informed decisions.

Conclusion

In conclusion, our podcast episode on AI and Pricing shed light on the complexities, challenges, and opportunities associated with leveraging AI in pricing strategies. While predictive models have their limitations, rule-based approaches offer control and adaptability in dynamic markets. The unpredictability of disasters and the need for explainability emphasised the importance of understanding the causal links between pricing decisions and market dynamics. We discussed the suitability of AI based on business scale and ROI considerations, highlighting the value of human expertise in conjunction with AI. By adopting a stepwise implementation approach, businesses can gradually incorporate AI into their pricing strategies, starting with basic automation and progressing to more advanced techniques. Ultimately, the goal is to optimise pricing decisions while maintaining transparency, trust, and a deep understanding of the market.

 

Related Podcasts