Today there is more data available to organizations, which has created a growing appetite to uncover valuable insights. As a result, easy-to-use software for predictive analytics is becoming more prevalent to help identify trends, solve problems, and uncover opportunities. PBCS Predictive Planning is Oracle’s way of providing intuitive predictive analytics tool, built right into PBCS.
What is Predictive Analytics?
For some, predictive analytics may seem like a relatively new concept. However, it has been around for some time now. It has become increasingly more popular in the last couple of years due to advances in technology, such as machine learning and big data, the need for competitive advantage, and growing volumes of data at organization’s disposal.
Predictive analytics is the use of historical data and statistical algorithms to identify the probability of future outcomes. For example, mathematical models use historical data, which takes into consideration trends and patterns in the data and applies the model to the current data to predict the likelihood of future outcomes. The information that predictive analytics provides helps organizations identify opportunities for positive change. Analysts use predictive analytics to reduce risks, improve operations, and ultimately increase their bottom line.
The Value of Predictive Analytics
Previously, business owners and analysts have relied on financial statements and reports to inform their decisions for the future. While this is valuable information, it limits to only telling us the past and why it happens. To stay ahead of the competition, organizations must be able to identify opportunities and pain points quickly. Additionally, this is where predictive analytics can be used in a variety of ways to provide value to an organization. Below outlines sample applications:
- Sales Forecasting
- Trends and patterns to predict revenue based on sales pipeline
- Customer Retention
- Details from paying customers to determine how likely they are to make additional purchases
- Product Pricing
- Trends and patterns to identify the seasonality of a product, what price it sells best at, and how it did at a ‘sale’ price
- Optimization of Marketing
- Insights such as social media activity, website clicks, email campaign results, and purchase patterns to determine leads and product market
- Inventory
- Trends and patterns to predict demand
- Employee Retention
- Employee details to determine most likely to leave, based on performance, pay, tenure, attendance and socio-demographic
- Accuracy of Budget or Forecasting Data
- Trends and patterns of historical data to determine future values and likelihood to hit given goals
PBCS Predictive Planning
Predictive Planning is Oracle’s predictive modeling tool and is a feature in Oracle Planning and Budgeting Cloud Service (PBCS). It was previously only offered as an extension within Oracle’s Smart View Excel add-in, but has been simplified and integrated with the planning forms. The predictive planning feature can be viewed in the planning forms within Smart View or the web.
Provided your Planning form is set up appropriately for Predictive Planning (need a Time and Non-Time Axis), a chart will then be displayed with trend lines relative to actual and current forecast data.
The model uses the historical data to provide a predicted, best-case and worst-case range. There are many options to customize the prediction models. Options include seasonal and non-seasonal models, with seasonal models highlighting (in blue) bands of seasonality in the historical data on the chart. There is also the option to choose the error measure for the model to use. If you do not specifically choose a prediction method or error measure, Predictive Planning will use the most accurate method based on your dataset.
Below are the list of prediction methods and error measures:
Non-Seasonal Methods
- Single Moving Average
- Double Moving Average
- Single Exponential Smoothing
- Double Exponential Smoothing
Seasonal Methods
- Additive
- Multiplicative
- Holt-Winters’ Additive
- Holt-Winters’ Multiplicative
Time Series Forecasting
- ARIMA = Autoregressive Integrated Moving Average
Error measures
- RMSE = Root mean squared error
- MAD = Mean Absolute deviation
- MAPE = Mean Absolute percentage error
Similarly to any other predictive analytics models, the more historical data that the model can use, the more accurate the prediction.
Finally, predictive planning allows you to manually or automatically copy the model prediction to your working forecast version.
Why use PBCS Predictive Planning?
Forecasting is necessary for all organizations. For instance, predictive planning can assist with the accuracy of said forecasting by predicting future values based on historical performance, trends, and patterns. Furthermore, predictive planning can give you an added advantage to increase your bottom line.
This easy-to-use feature is available for download in the downloads section of your PBCS pod.
Download our eBook, “Turn Waste into Value”, to learn how to find true value in the planning process, and 3 steps to easily get your company started on the path to effective annual planning.
