Time series analysis is a standard form of analysis that’s used quite often. Time series analysis takes an in-depth look at time series data, which is data that changes over time or for which time is considered a variable in the results. Time isn’t just a measurement included in the data—it’s the primary axis on which the data sits.
The main difference for time series analysis versus other analysis is that the data is collected over regular intervals of time. This helps the analysis identify systemic patterns in the data that help form trends, cycles, or seasonal variances. When there is a consistent time frame of historical data, forecasting can be used to predict likely future data.
Because time is an integral variable in data for many industries, a variety of applications for time series analysis exist. Explore a few time series analysis examples below.
1. Exelon uses data analysis for data-driven audits
Problem: Traditional audits were time-consuming and did not add much value.
Solution: Exelon needed to conduct regular audits of financial processes, and traditional audits take time. Audits used to involve interviewing the counselors or administrators who own the processes and see the records at the time of the audit, but Exelon wanted to do more. The company used Tableau to look at an entire year of data and uncovered trends they may have missed.
Stock prices are also reported over time and involve time series analysis. Trading algorithms that work autonomously also utilize time series analysis, collecting data on the ever-changing market to react to minute financial changes and trade appropriately. In other areas of finance, time series analysis can be found in trend analysis, seasonality, budget analysis, interest rates, sales forecasting, and financial markets. Since finances are so regularly recorded, it makes it a suitable subject for analysis over time.
Beware that time series analysis of finance data can include so many variations that complex models are required. A model that is too complex can lead to either lack of fit or overfitting, which do not distinguish between noisey errors and true relationships, resulting in skewed analysis.
2. Stamford Health uses data analytics for patient care and reducing costs
Company: Stamford Health
Problem: Inefficient use of resources and inflated costs of care and operations over time.
Solution: Stamford Health used data analytics to identify opportunities to improve patient care and reduce costs for patients and the system. The company used the historical length of patient stays, treatments, and conditions data to chart when patients received certain treatments and how that affected patient outcomes. Using these combined data sources and data analysis, Stamford Health identified better times to administer medication and reduces the average length of stay. This reduced both patients and hospital costs.
Healthcare professionals have been making great strides with data, both through patient care and technological advancements. While informatics improves patient care and patient information, and the Internet of Medical Things automates and augments patient data — time series analysis is found in chronic disease research. With time series analysis, chronic diseases, defined as diseases that last a year or more and require ongoing medical attention, can be tracked over time, as time is a major component of these diseases. In the same vein, time series analysis plays a crucial role in epidemic-scale research.
For everyday healthcare, time series analysis is used to monitor a patient’s heart rate through an electrocardiogram (EKG) or even brain waves through an electroencephalogram (EEG). The devices record electrical signals coming from the brain or heart over a period of time, identifying abnormal activity that indicates a medical issue.
3. Texas Rangers use data to identify sales opportunities
Company: Texas Rangers
Problem: Data analysis was not fast enough to make decisions days before game day.
Solution: The Texas Rangers front-office team combined all their data sources so they quickly had a 360-degree view of the data. With advanced dashboards, sales and marketing teams used time series analysis and other data analytics strategies to quickly identify opportunities, especially related to forecasting against seasonal trends.
In one practical example, the sales team looked at up-to-date dashboards and realized that their projected sales for an upcoming game were lower than normal. About a week before the game, the marketing team strategized on how they could increase ticket sales. They developed a marketing strategy four days before game day, but they had time to create a promotional Father’s Day ticket offer to increase sales.
4. Bronto Skylift uses more accurate forecasting to save costs
Company: Bronto Skylift
Problem: Operations, manufacturing, and sales forecasting was inaccurate and time-consuming.
Solution: Using better data and faster analysis, Bronto Skylift cut analysis time from one day to one hour. Using time series analysis and forecasting modeling, the company can forecast supply chain and processes in its manufacturing department and forecast seasonal trends. Before investing in data analytics, the data was stale and siloed.
Now these forecasts are much more accurate, reducing costs in inventory, supply chain, labor, and capital equipment.
Begin your own application of time series analysis
Because time is an essential variable in so many data models, time series analysis has a range of applications—only a few of which are listed above. The Tableau platform provides comprehensive time series analysis with the built-in date and time functions that allow you to drag and drop to analyze time trends and easily perform time comparisons, like year-over-year growth and moving averages.
Begin your own application of time series analysis with an easy-to-use visualization software to easily identify trends, find outliers, and compare data over time. Check out more time series resources and customer stories to help you get started.