AI Enabled Price Elasticity for the win!

Why did the demand curve go to therapy?

Because it couldn’t handle all the emotional highs and lows—its price elasticity demand was just too unstable!

Sorry for the bad-ish joke! It’s not original – It’s GPT written. But still, portrays the fact about the confusions, the ups, and the downs that need to be considered while talking about elasticity of pricing.

In our previous blog about pricing analytics, though we’ve spoken about elasticities – but it was like scratching the surface of an iceberg. So, this article is to talk more in depth about Price Elasticity of demand and to understand why it is a tricky mess.

The simplest way to understand price elasticity

Price elasticity is the attempt to reduce demand changes to a single unit. That’s it. The longer version is that for calculating the price elasticity you need to divide the change in demand for a product (can be a service, resource, or commodity) by the change in price.

Formula for calculating the price elasticity demand

Price Elasticity = Percentage Change in Demand of Quantity / Percentage change in price

Or something like:

Price Elasticity = ((New Quantity- Initial Quantity) / (New Quantity + Initial Quantity) / 2) / ((New Price – Initial Price) / (New Price + Initial Price) / 2)

The result of this is usually a negative number (given that quantity & price have opposite effects i.e. increased quantity for decreased price and vice-a-versa) which lies in between 0 and 1.

Wish it was this simple to represent it anyways. The above linear graph is usually represented when we deal with one product. But unfortunately, the real world doesn’t have just one product and neither a pricing that’s this simple. Remember, the final price includes the promotion differences too. So, how’s it affected?

Factors affecting the price elasticity of demand

Imagine that you’re buying Milk – the closest Walmart would have options, ranging from Whole Gallon, Vitamin D infused, Skimmed, Low fat and more. In the choice of what to buy, how does the elasticity play a role? Let’s get into the details.

1. The type of product

Our example here is milk, which is a staple in most houses and has easily available alternatives. More substitutes are available when more elastic the product is, as customers tend to switch. What if they were looking for an electric toothbrush, given the lesser variants available in the market, the switch would not be easier i.e. with inelastic goods. AI techniques would be useful in identifying the parameters that customers perceive to the product to be a necessity vs luxury and the weightage of each of these factors – to decide on the elasticity.

2. Base price impact on Revenue

Let’s go back to the graph above: The elasticity to go from point A to point B is different from going from point C to point B – the change can be due to a perceived price threshold – depending on the original high or low price – the lower the OG price is – the inelastic it might be. (Depends on the product)

In the Revenue vs Price plot for an inelastic product– you can see that the increase in price would always result in increased revenue (revenue gained due to increase in price) is more lost due to decrease in volume. Whereas for an elastic product, the decrease in price (even with promotions) would result in increased revenue.

3. The effect of Promotional elasticity

It is ultimately the price that the consumer pays that would be considered as the final price. This is the Halloween season (while writing this), you open Walmart and see the three options below for buying trees? Though we’ll not go into the obvious choices – but the impact that promotions play on decision-making is huge.

And this is just one channel. In your overall promotional strategy, it is important to the find the right impact of the promotions across different channels. This for example, is a snapshot of a part of the trade promotion section of our RGM tool – which helps users understand the impact of their various promotional activities.

Extra note: The importance of finding the right price format

Everyone loves promotions and thinking that they’ve got a good deal – that’s the end goal! To make this happen you need to identify the actual elasticities and their movement. So, you’ll need incorporate factors like promotions, seasonality, and competitor pricing. Then classify the products with Support vector machines, Gradient boosting, etc.

You can use hierarchical modelling to handle complex relationships, and more. The graph below summarizes the relationship between base price and promotional price effectiveness!

4. The competitor conundrum

Let’s be honest, we’re not living in an idealistic world – the price elasticity is greatly dependent on how the competitors are changing the prices. Here is a simulation on how the price elasticity would change based on the promotional and competitor pricing.

5. Is it noise or the data?

Given all these conditions, you never know the reasons behind some buys – and it is needed to clear them out before you start with the analysis. Someone saw Taylor Swift wearing something similar or there was a game for which snacks were needed or it was a long weekend and people were on vacation.

What’s the noise? In addition to random noises like this – there might be systemic noise like things that are bought every August, or something that people buy when the temperature was too high?

With AI and analytics techniques like Anomaly detection and signal processing, you can help filter out market noise and identify genuine trends in price elasticity.

There’s one other glaringly missing thing we’ve avoided talking about – having the right tool. You might know the fact that the competitor data is important or that adding promotions is important – but you might lack the tool to give you the desired results by taking in the multiple parameters and attributes into account!

AI-Enhanced Price Elasticity Strategy for RGM

We’ve spoken about the complexities of building the right price elasticity model till now. This gap can be solved with the help of AI & analytics techniques. But the impact is not very easy to understand especially with techniques and methods like price elasticity. Also, this is not something ChatGPT or Perplexity can help you a lot with.

One important thing to note before we get into the AI models and improving the data for RGM is to consider the granularity of the elasticity you need. This is quite dependent on the granularity of the data available with you for the analysis of price elasticity.

Here a few ways in which AI is being used to measure Price Elasticity:

• Linear models like multiple linear regression and regularized versions (Ridge, Lasso) are simple but may not capture complex relationships.


• Non-linear models such as polynomial regression, SVR, and neural networks can model more complex elasticity patterns.


• Bayesian models incorporate prior beliefs and provide uncertainty estimates.


• Graphical models like Graphical Lasso and Bayesian Networks help understand relationships between variables.


• Ensemble methods (Random Forests, Gradient Boosting) can capture non-linear relationships and variable interactions.


• Time series models are useful for dynamic, time-varying elasticity scenarios.

Now to the topic we skipped above: Do you have the right tools to calculate this?

Finding the right solution for Revenue Management!

Before we answer this question – there is a multi-faceted approach to the answer – you might be at a place where you’re relying on spreadsheets, or you might have descriptive and diagnostic analysis with dashboards for historical data, or you might have some level of AI enabled RGM for your brand. Though the starting point might differ – your goal would be the same – Seamless revenue growth.

Profit Pulse. AI: Our Explainable AI enabled RGM platform has linear, non-linear, Bayesian, and graphical formats of measuring your price elasticity – this is to ensure that the elasticity is being calculated with all the right attributes.

Our Revenue Engine consists of Marketing simulation to analyze the non-linear impact and uplift due to trade schemes and promotions. It also consists of Explainable forecasting and planning modules with a micro-geography and outlet level product recommendations.

Let us know if you’d like to discuss any of this in detail. We’d love to hear your feedback and thoughts.