Historical Analytics

Historical Analytics

When most people think of business analytics, it is usually Historical Analytics they have in mind.  As the availability of “big data” has grown, the use of analytics to gain insight into the data has increased dramatically.

The traditional definition of historical analytics

Historical analytics refers to the analysis of activity and data from the past to discern particular trends, patterns, correlations, and other statistical relationships that may drive insight into business performance.


How Historical Analytics Relate to IoT Analytics

In the context of the overall field of IoT Analytics, historical analytics provides the baseline of past behavior and data for building up the more refined data and information necessary to serve the advanced and complex challenge of IoT use cases.  While surely necessary for taking advantage of the Internet of Things (IoT), it is only the first step in the multiple step process needed to leverage the full value of IoT.  To learn more about this read about the Analytics Value Chain.


Business Use Case Examples for Historical Analytics

In the usual process of historical analytics, typical questions and analyses might include:

  • What is the series of events that led to a particular outcome on a business metric?
  • What is the average number on a particular business variable or output?
  • How many total events of a particular type occurred in a particular time period?
  • What is the total range of responses or output in response to a series of data inputs?
  • How does the result of a particular process vary based on the speed of various sub-processes?
  • Where is the correlation in various measures or outputs? Causality?

In all these examples, the basic objective is to understand the past and what kinds of results happen in various scenarios.  And – when possible – to understand the causes of various outcomes.


Commonly Used Subcategories of Historical Analytics

Historical analytics is sometimes divided up into the sub-categories of descriptive and diagnostic analytics.  Descriptive analytics tries to explain WHAT is happening.  Diagnostic analytics is focused on explaining WHY it happened.

Examples of Diagnostic Analytics and Descriptive Analytics

With descriptive analytics, also known as data mining, the focus is to gain insight from historical data using reporting, score cards, dashboards, clustering, etc.  The patterns that are uncovered are useful for scenarios such as assessing credit risk where past financial performance data is used to predict a customer’s likely financial performance.  Or in retail, patterns can be used to categorize customers based on their likely product preferences and sales cycle.

With diagnostic analytics the focus is to look at past performance to understand why certain things happened.  For a social media marketing campaign, you can use descriptive analytics to assess the number of posts, mentions, followers, fans, page views, reviews, pins, etc. There can be thousands of online mentions that can be distilled into a single view to see what worked in your past campaigns and what didn’t. The result of the analysis is usually displayed on a dashboard.


Historical Analytics are Important to IoT Analytics

Historical analytics is one of the key components in the overall process required for IoT Analytics.  When coupled with real-time streaming analytics, predictive and prescriptive analytics are more meaningful.  The key to effective IoT Analytics is to combine historical and real time analytics flowing from “things” to drive better prescriptive analytics and enable Intelligent Actions.


Learn more

Learn more about how IoT Analytics are: