Using Machine Learning For Pricing Optimization in a B2B Setting

B2B sellers can charge their customers different prices for the same products, and often can even change the price from one purchase to another with little notice. Obviously, this kind of “price discrimination” creates plenty of opportunities to increase short-term profit. However, it can negatively impact the development of trust needed to establish a long-term relationship with buyers. For this reason, a good statistical methodology should be able to predict the short-term impact of pricing as well as the long term effects of pricing decisions.

B2B pricing decisions differ substantially from those within the business-to-consumer (B2C) market because of the high complexity of the business transactions, where the customer needs to make several-interrelated decisions, and the salesperson needs to be able to “read” the customer mindset before generating a quote.

At Dunn Solutions, we realized that it is this complexity that creates tremendous potential for increasing profits by leveraging machine learning.

This video shows how machine learning can be leveraged to accurately model all these interdependent decisions and maximize profit in B2B!

For more information, please visit: www.dunnsolutions.com.

Video commentary:

"Let’s talk about Machine Learning in a B2B setting. The B2B sector is a huge part of the US economy, but it has been very slow at adopting Predictive Analytics. There is a misconception that because B2B transactions are so complex and the human component is so important that predictive analytics is not as useful.

I take a different stand: it is this very complexity that creates a lot of opportunities to leverage Machine Learning/Predictive analytics. One area of interest, is Dynamic Target Pricing, where the complexity comes from having to simulate all the decisions the buyer makes and how each decision has an impact on the very next one.

What is Dynamic Target Pricing? We need to understand that pricing decisions in B2B are much different than those in B2C. For one thing, sellers can charge their customers different prices for the same products. For example, customer A might receive a completely different price for Product A than Customer B does for the same product.

Also, the seller can usually change the price with little or no notice. Now, obviously this is a kind of “Price Discrimination” and creates many opportunities to make a quick profit in the short term. However, it can also negatively impact the relationship the seller has with the buyer, and affect the long term profit. Therefore, if you want to use machine learning for pricing you need to make sure you design a methodology that accurately predicts not only the short term but also the long term effect of your pricing decisions.

In the next part of this presentation, we will present a case study to show how we created such a model using transactional data from one of our B2B clients. But before we do that...

There are a few concepts we should be familiar with if we want to use Machine Learning in a B2B setting:

• First, the business of selling commodities or manufactured goods (or both). And the reason why commodities are homogenous in nature and tend to be more price sensitive. Let’s say you are a B2B retailer of lumber. Lumber offers little to no differentiation, so your buyers will pay more attention to the prices of your competitors.

• External Reference Price: This is really more important for commodities; but again, the price of commodities is easily available and your buyers can easily reference it to determine if they are getting a good deal.

• Internal Reference Price: this is the individual-specific price that your buyers have paid in the past. Buyers tend to reference the price they were charged in the previous transactions to determine if they are getting a good deal.

• The effect of a price increase is not necessarily proportional to the effect of a price decrease. In other words, if lowering the price by 1% increases sales by 2%, it doesn’t mean that increasing the price by the same 1% will decrease your sales by 2%.

I do not suggest in any way that a Pricing Modeling should or even can replace the salesperson; but I do see it as a tool to equip the salesperson with that can generate profits significantly higher than the salesperson alone. But we have seen that creating a robust pricing model is challenging. Please don’t hesitate to contact us for a Free-of-charge assessment of your data and we will gladly tell you if and how you can leverage it to increase your profits.

I want to thank you all for your attention, and I hope to see you for the next webinar!"

Видео Using Machine Learning For Pricing Optimization in a B2B Setting канала DunnSolutions