Today, setting the price of products and services can be tough enough, without taking into account a constantly shifting marketplace. Competitive prices, product releases, inventory levels, trends and seasons are all crucial factors when considering smart pricing strategies in an agile marketplace.
The question is, how do we develop smart pricing strategies that take account of all these factors? How can we gain better insights into developing and implementing our pricing strategies and influencing our bottom line?
Using an advanced analytics approach to price sensitivity analysis that reaches far beyond the “classical” measurements of price elasticity, is an important first step in attaining the required insights for smart pricing strategies.
Being able to accurately predict price sensitivity may seem far-fetched, but it is in fact a fast-approaching method using data-driven analyses and machine learning that few industries have (so far) contemplated. For companies in retail, banking, entertainment and media, there can be opportunities to learn more about customers and pricing through data that may already be available in-house, paving the way to actionable insights, increased profits and staying one step ahead of competitors.
It is clear that pricing will become much more automatic in the future. Today, companies are sitting on masses of data that is not being leveraged for pricing purposes. With the right models in place however, companies can pinpoint what price their products and services should be at the right time to maximize profit, revenue or competitiveness.
Let’s imagine that last week you sold 1000 units of a computer for $799. This week, you want to examine what happens to sales if you reduce your price to $749. The classical approach would be to estimate the price elasticity of demand for that product by analyzing the historic development in the relationship between price and sales. The problem with this approach is that there are more factors than just price that affect sales. For instance, the product may have been included in a large campaign, competitors may have lowered their prices, or a certain market trend might have caused an increase in popularity of that product. This is where machine learning and data becomes a benefit, because they allow us to analyze the complex relationship of how numerous market factors influence how our customers perceive prices. As a result, it enables us to predict how price sensitive they are, considering the current market conditions. In fact, predicting price sensitivity is closely related to predicting demand, and it goes without saying that having an accurate demand estimation model can be incredibly valuable.
Certain industries such as retail are already trialing predictive price sensitivity to help set smarter prices for their products. To begin the process, key variables that could affect market conditions, demand and price sensitivity are identified using advanced analytic methods. These variables can be identified within a company’s own data, as well as from external sources.
Once these variables are pinpointed, a tailored, price-sensitivity algorithm is created, tested and implemented within the retailer’s pricing infrastructure to provide automated predictions of the potential impact of a suggested price, taking into account the current market conditions. This reveals insights as to how customers might respond to changes in price, which, in turn, enables the retailer to close in on an optimal price.
Clarify what your pricing strategy - and your business - is trying to achieve and why. The answers will influence how you structure your data and implement your solution. However, if your pricing strategy is to consistently set the lowest prices, the potential gain from doing predictions on price sensitivity is strongly diminished. This is because regardless of what your customers’ price sensitivity is, according to your strategy your price will be dependent on your competitors and not your customers’. It is therefore advised that you only use this method when it aligns with a more varied pricing strategy.
Don’t underestimate the process of collecting, repairing or engineering the data already available to your business. The challenge in predicting price sensitivity is creating a data set that accurately represents the important variables and dimensions that could have something to say about your customers’ price sensitivity. To tackle this, reviewing data assets and competencies will clarify the relevant data sources you have, those you don’t need and those that still need to be gathered.
If your business doesn’t have a strategy for capturing data, consider establishing one. If it does, consider how best to collate and analyze it in the future. A gathered history of data is vital for an effective and well-tailored price sensitivity algorithm.
Examine what the minimal viable product is for your business in terms of price insights. Then start building more advanced solutions, incorporating new and aligned data sources, and state-of-the-art algorithms. By starting small, the road to usable results gets shorter, and ultimately, with more data and experience comes the potential to increase predictability and the degree of automation.
Price is one of the most important levers a company can use to boost its profits. According to a famous McKinsey analysis, a one percent increase in price will, on average, cause an eight percent increase in operating profits for an S&P 1500 company, assuming sales stays constant after the price increase.* The fallacy here is to assume that this is true for all products, when it is in fact highly dependent on the customers’ price sensitivity. However, it exemplifies the power of pricing, and by predicting price sensitivity, we can identify the products in which there is a potential to increase prices without causing a negative effect on sales. In another recent study, 78 percent of retail consumers said it was reasonable to use data science to increase and decrease prices, as long as it was a price they were prepared to pay**.
As your model accuracy and experience of predicting price sensitivity increases, so will your ability to use these models to develop fully-automated, dynamic pricing algorithms. This will position your business to outsource the pricing process to an algorithm that can alter the prices of your products in real time, and account for the relevant market factors and demand. You’ll then be closing in on offering the optimal price at any given time.
Remaining competitive and retaining customers has never been more challenging, and the need to adapt to a constantly changing marketplace calls for fresh perspectives and new ways of working. Reviewing and aligning your business’ current data sources can define the scope of your future pricing policy. In addition, it can produce actionable insights on how the market will respond to the steps you take, and capture correlations between the different variables of the market.
Across industries, predicting price sensitivity will only become more prevalent. Implementing this suggested three-step approach is a practical way of identifying wider, long-term opportunities, so start now and stay ahead!
* Source : https://www.mckinsey.com/business-functions/marketing-and-sales/our-insights/the-power-of-pricing
** Source: http://pages.revionics.com/Forrester-Shopper-Report.html