How to define your pricing strategy
Pricing across multiple markets
AI is undoubtedly the ultimate goal for optimal pricing. If artificial intelligence can identify causal relationships between variables, it can significantly improve pricing optimization. However, beneath the hype and appeal of AI, there are several limitations to consider.
AI will never fully explain why it makes certain decisions, meaning you lose some control over your pricing strategy. Additionally, for AI to be truly effective, it requires a massive dataset. If you don’t process tens of thousands of orders per month, AI-driven pricing may not be a viable option. If you do consider AI, be sure to question your provider thoroughly about how they measure and evaluate data to avoid misleading or inaccurate results—AI models are often prone to false positives.
For fully automated repricing, algorithmic models can work well, but they require an enormous amount of data to be accurate. That’s why they are primarily used in FMCG—if you sell 600 packs of butter per day, AI will always outperform an experienced product manager. However, this level of sales volume is rarely found in e-commerce.
Regardless, we strongly recommend storing as much sales data as possible, so you can later analyze seasonal trends and refine your pricing approach.
There are also heuristic-based models that are more suitable for e-commerce. Put simply, these models increase prices when products sell well and lower them when sales slow down. However, even with these models, always ask your provider about the data sample size and methodology. Treat AI as one of many tactics to test, not an automatic silver bullet.
Ultimately, AI models follow the same principle as other pricing tactics—they are just one piece of the strategic pricing puzzle. Like any other tactic, they should be tested, analyzed, and evaluated. The next chapter of Disivo Pricing Academy covers exactly how to assess pricing tactics effectively.
