
Guiding Principles Of Developing And Narrating Analytics Insights
-
Identify template for each of the data models (Incident, Problem, Change, Service Requests, RITMs, Event Management, Application Transaction data, etc.)
-
Identify data source and validate query to retrieve the data
-
Perform QA on the retrieved data to ensure alignment with template and query
-
Transform data using ETL processes for analytics preparation
-
Harvest the data to capture historical and forward-looking insights
-
Start with a Quantitative Analysis.
-
Look at Process Behaviour Analysis (PBA) Charts (Opened Volume, Closed Volume, Response Time in Mins, Resolve Time in Hrs) from a Time-Series perspective (By Year, By Month, By Week, By Day, By Hour) and see if you detect any signals, patterns, behaviours, abnormal behaviours, anomalies, etc.
-
Point to Point Compare: How is the current data point compared to the previous data point? Is it higher or lower? If so, why? What is causing the increase or decrease?
-
Check current period Category list to previous period
-
Check current period SubCategory list to previous period
-
Check current period Configuration Item list to previous period
-
Check current period Requested Item list to previous period
-
Check current period Resolution Code list to previous period
-
Check current period Close Code list to previous period
-
Check current period Contact Type list to previous period
-
Check current period Affected User list to previous period
-
Check current period Organization list to previous period
-
Check current period Company list to previous period
-
Check current period Country list to previous period
-
Check current period Location list to previous period
-
Check current period Business Functional Unit list to previous period
-
Check current period Assignment Group list to previous period
-
Check current period Priority list to previous period
-
Check current period Short Description list to previous period
-
Check current period Incident State list to previous period
-
Check current period Category list to previous period
-
-
Exceptional Behaviour: Are there points above or below the Upper Control Limit
-
Exceptional Behaviour: Are there points above or below the Lower Control Limit
-
Shift in Process: Are there 8 consecutive points (8 consecutive points means a change has occurred and been established) above the average. However, at 7 points in a row, you should start to be more cognizant, at 6 points in a row, you need to start paying attention.
-
Shift in Process: Are there 8 consecutive points below the average.
-
Trend in Process: Trend up (6 points, all increasing from the previous point).
-
Trend in Process: Trend down (6 points, all decreasing from the previous point).
-
A regularly reoccurring low number: Determine if there is any seasonality to the data.
-
A regularly reoccurring high number: Determine if there is any seasonality to the data.
-
Gradual Trend: Values that go up steadily over the entire chart, but never quite 6 in a row.
-
Saw Tooth Pattern: Data that alternates sides of the average for 14 points, (7 pairs of points).
-
Two of Three points outside the two sigma limit.
-
Four of Five points outside the one sigma limit.
-
Moving Range Chart: Where you go from very low value to a very high value, both within the limits, or you go from very high value to very low value, again both within the limits. This can identify this single change is excessive and worth investigating.
-
-
Look at the Pareto Charts and to see if you detect anything abnormal or changes since previous report. Look at Category, Subcategory, Root Cause, Contact Type, Assigned To, Assignment Group, Affected User, Business Unit, Organization, Location, Etc.
-
Look at the Matrix Function to see if you can detect changes from a MTM view of top categories (Subcategory, Assignment Group, etc.) compared to a baseline Month.
-
Look at the Compare Function to see if you detect changes in top categories when compare two date/time data points. MTM, WTM, DTD, and HTH.
-
Use the Custom Report Function to look at additional insights for further drill-downs.
-
-
Conduct a Qualitative Analysis for further Insights
-
Look at the Slice n Dice Function to see if you can further drill down into the observations you have made.
-
Use the Manual PBC Function to build custom PBA charts.
-
-
Tie the findings/observations to gaps to the Operational Management Reference Framework.
-
Convert the gaps to recommendations based on the Operational Management Reference Framework.
-
Transfer the recommendations to the CSI Register and manage to closure.
Notes:
-
Turn data-driven insights into powerful narratives or stories that would grab the attention of all stakeholders and compel them to listen
-
Narrative is the way we simplify and make sense of a complex world. It supplies context, insight, interpretation – all things that make data meaningful and analytics more relevant and interesting.
-
With analytics, your goal is normally to change how someone makes a decision or takes an action. You are attempting to persuade, inspire trust, and lead change with these powerful tools. No matter how impressive your analysis is, or how high-quality your data, you are not going to compel change unless the stakeholders for your work understand what you have done. That may require a visual story or a narrative, but it does require a story.
-
Most people can’t understand details of analytics, but they do want evidence of analysis and data. The most compelling stories of all are those that combine data and analytics, and a point of view.
-
Data preparation and analysis often take time, but we need shorthand representations of those activities for those how are spectators or beneficiaries of them. It would be time-consuming and boring to share all the details of a quantitative analysis with stakeholders. Analysts need to find a way to deliver the salient findings from analysis in a brief, snappy way. Stories fit the bill.
-
Couching the analytical activities in stories can help to standardize communications about them and spread results.
-
Use the data and analytics to find solutions to problems.
-
Use the findings to tell the most convincing and compelling story.
-
Storytelling is, in fact, the final and most important stage of analytics. Successful communication of findings to stakeholders is as important as conducting robust analytics.
-
The strength of the data lies in the power of the narrative
-
Hypothesis testing is an approach to determine if the difference in the observed data is real or merely a matter of chance
-
Graphics are uniquely suited to present the interplay between several variables
-
Build a strong narrative from your empirical findings and then communicate to the stakeholders
-
Use the analytics to have fact-based discussions
-
Successful communications to stakeholders is as important as conducting robust analytics
-
Use the narratives to strengthen an analytics-drive decision making process
-
Most analysts are poor storytellers. They are introverts, more comfortable with machines and numbers than with humans. They are taught to focus on empirical disciplines.