Price optimization is the process of using data to determine the optimal price of a product, based on an objective or set of objectives. Hmmm…sounds pretty boring when you put it that way. But when you put it to work in your business, it gets a lot more interesting!
To set prices, you need to have your goals in mind. Do you want to dominate market share? Do you want to get the highest possible margin? Do you want to maximize profit, but sell at least 10,000 units to get rid of all your inventory? Whatever objective(s) you set, you can use data and algorithms to determine the optimal selling price (or prices, depending on customer/market/timing) to meet that goal.
It all starts with your data
Let’s take a step back, though. Before you can do the advanced predictive analytics of price optimization, you have to start with good, clean data. This data can reveal how much flexibility you have in your pricing before customers change their buying behavior (commonly known as price elasticity). You can find this out by studying your historical data to see what happened last time you changed prices. You might find that Customer X bought a lot less, while Customer Y bought slightly less, and Customer Z kept buying at the same pace. Of course, we want software to do this tedious work for us.
Knowing the behavior model for each customer helps you determine what price to give each customer to help attain your stated goal(s). Through many, many iterations, you can input possible prices, the customer behavior model, your goals (and their relative priority) as well as constraints (limited inventory or production, minimum/maximum prices relative to competitors, etc.) into the algorithms and measure predicted outcomes to try to achieve as many goals as possible. This price optimization based on goals and constraints is very complex to perform manually, and you cannot scale this process for the number of customers and products you have without automation tools.
To make the pricing process more manageable, you should group your customers based on how sensitive they are to price changes. This allows you to price several or many customers with similar price elasticity behavior at once, saving time in the pricing process. Again, this grouping exercise (known as segmentation) is best done by software because of the intensive effort to define and evaluate what “similar” means and to determine what attributes these similar customers have in common.
Why do you care about customer segmentation through similar attributes?
Because when you encounter a new customer with no history of price-change reactions, you still need to give them a price—the “right” price. How do you do this? You try to predict how this customer might respond to a price by comparing him to similar customers with known history. Software can match new customers with an existing group based on the common attributes of the customers already in the group. And this group already has an optimal price that can be suggested for this new customer as well. Thus, price optimization not only helps you price existing customers but also to maximize profit on new customers.
The best part about price optimization is that, once you get it up and running, it can very quickly pay for itself by finding opportunities to make more money on the deals you’re already doing. The software keeps you from leaving money on the table every time you make a deal. And all this extra money is pure profit. No additional cost of sales, no extra promotions, etc. You’re simply adjusting prices with very little impact to sales, but a huge impact to profits.
The good news is that data science isn’t rocket science.
Well, it’s close, but companies like Vistex now offer “data science in a box,” so you don’t have to be a rocket—I mean data—scientist to use data science in your everyday pricing processes. That means smaller companies can now get the capabilities and advantages of data-powered pricing that were once only available to the companies with the deepest pockets.
To learn more about analytics and pricing, view our webinar, How to Win with Analytics.