Best-in-class fuel retailers have a few things in common, but one of the most prominent traits is a commitment to using data for decision making. Without strong data, making truly informed, strategic fuel pricing decisions is next to impossible. That's true with regard to any of the 7 Elements for Fuel and Convenience Retail Success, but especially true in fuel pricing. Though experience and intuition can play a part in a great pricing strategist's approach, nothing beats data.
But what kind of data? And what role should it play in your fuel pricing strategy?
Various data types must be taken into account when developing a fuel pricing strategy. Some key data points include:
- Planned volumes (forecasted)
- Sales volumes (actuals)
- Own prices (historical and current)
- Delivery volumes to site
- Price requests and price approvals (historical)
- Wet stock volume
- Market share
- Traffic data
- Events (recurring and ad hoc)
- Location, facility, operations, merchandising, and brand data
With information about each of these topics on hand, you can begin to construct both strategy and tactics that work for your specific site or retail network, bringing you ever nearer to a higher level of predictability. But it's not just the data itself that's valuable. In order to join the ranks of best-in-class retailers, you also need to collect the data at the right times, access it at the right frequency, and ensure it's clean for modeling.
When is the right time to collect data? In short, as regularly and granularly as possible. Regularity gives you trends. Granularity allows you to drill down into those trends. Most of the data you input into your analytics and model must be as up-to-date as you can manage, lest it be rendered useless due to inaccuracies created by even subtle market shifts. This recentness is also important to allow you to continually monitor, evaluate and improve your fuel pricing processes.
Of course, updating data too frequently can also have its drawbacks, including that doing so may encourage you to update price too frequently, contributing to an unhealthy level of volatility in your surrounding market. In your quest to gather the right data at the right time, do not leave out consideration of seasonality. Some data only changes with events or time periods and won't need to be updated with every new modeling activity.
In sum, because all data points require different treatment, understanding purpose is paramount. You must understand the purpose of the data in order to understand when to collect it. Data won't need to be recent or complete if, for example, you are using it only to directionally inform your perspective on seasonality in order to assess your medium-term pricing strategy. However, if you are using data to inform daily tactical decisions, recentness and completeness are especially important.
Data's role in shaping a fuel pricing strategy is virtually unlimited. It informs and allows you to create insights which drive short, medium, and long term strategy, and create efficiencies to streamline the pricing process. But that's only true of the right data, gathered in the right frequency. So, how do you get access to that data? What is its availability?
You may find that you require additional data suppliers in order to create the level of insight needed to enhance your pricing strategy demands. To discover which data will produce the greatest benefit to your pricing model and thus have the strongest cost benefit analysis, you'll need to have a strong understanding of which data points are the most predictive. Do traffic counts typically contribute to your volume and therefore, pricing changes? Is an event the most important contributor to demand? Is it demographics? Competitor information? Something else? Discovering which data is most predictive is a sophisticated process, but one that's necessary, if you are to see the level of success all pricing strategists dream of.
We have touched on the various data points fuel pricing strategies should include, the timing and accuracy of those data points, and their value in creating predictability. But consistency is also critical for maintaining your data integrity — and the integrity of any modeling activity or system you choose to leverage.
Before inputting any data into your model, a transformation must occur. The data must be clean and well-structured to your purpose with outlying and egregious data points removed. When global data comes into play, this need is especially apparent. When taking data from numerous countries and using it in combination with factors such as privacy regulations, language, and more can create major data challenges in understanding commonality across disparate formats.
Data can shape your fuel pricing strategy and every tactic that strategy requires to be executed at your sites. Its role, as mentioned, is virtually unlimited. It can allow you to predict the future with a known degree of certainty, so you can better forecast volume and therefore make stronger decisions about the rest of your sites' components. It can allow you to account for your operational realities, including the accuracy of your past prices and how often those have been rejected or approved. Perhaps most importantly, when modeled and leveraged correctly, data can push you into a new way of thinking about fuel pricing — a way the best-in-class fuel retailers already understand.
To learn more about best practice data usage in fuel pricing, join Kalibrate at our San Diego User Conference in March. Register here.