Contribution analysis is a product feature that helps you understand your data better and identify the root cause of unexpected anomalies in detail.
With our continuous pursuit to help businesses reduce the blind spots in business metrics, we have come up with yet another feature that helps you find the contribution of all the factors that led to an anomaly. The primary goal with Contribution Analytics is to provide the user with Automatic Diagnostics and help the user in root cause analysis.
Let me walk you through an example to illustrate the power of Contribution Analysis. Consider a scenario where a company’s revenue has a sudden dip. With Anomaly detection turned on, the dip in revenue is notified to all the stakeholders in the organisation. But to the stakeholders, the story is still incomplete. Only an anomaly at a particular time does not paint the whole picture as they still have to investigate further to determine the cause of the Anomaly. If any incident has caused it, then they have resolved the incident to prevent any loss to the business.
With Contribution analysis, we are trying to address the manual task of investigation done by stakeholders to identify the cause of the anomaly through Auto Diagnostics. The Contribution analysis automatically finds the factors that may have caused the anomaly and shows the top factors. The Stakeholders will now have a better picture of the anomaly as they have the notification and the top factors that might be causing the Anomaly. It helps immensely by cutting down the effort and the time is taken to investigate the cause of the Anomaly and providing near real-time insights to the user to take corrective action.
Local and Global Context
Contribution Analysis has two contexts in the CrunchMetrics application as follows:
Local Context: At a particular Anomalous point, it shows all the factors that may have contributed to the cause of Anomaly along with the percentage of contribution. It helps in getting the immediate context of factors for each anomaly.
Global Context: During a time range, the KPI would have varied by a certain percentage. The Global Context would generate a report which shows all the factors that may have contributed to the Variation in the original metric as well as the variation of the factors with time.
In this case, since we are looking at the variation of the metric within a Time range, our focus on the factors contributing to the variation in the metric is obvious but also it is critical to understand that the factors also change with time. Therefore, keeping this in context we are also helping users with the Variation percentage of the factors along with the contribution percentage.
Power of Multivariate under the hood
Leveraging historical data, our proprietary AI/ML algorithm will automatically select all the predictors which can accurately predict the variation of metric as a function of other variables (factors). It evaluates the effect of other variables and indicates the order and magnitude (percentage relative contribution) of the effect that each of the other variables is having on the metric at any instance (anomalous point or series of anomalous points).
It is an unsupervised ensemble model which automatically selects the best fit model and automatically tunes the hyperparameter to get the best accuracy.
To see the Contribution analysis in action or to know more about it, please feel free to reach out to us.
Summon the power of Augmented Analytics to identify risks and business incidents in real-time.
Kshitish Sahoo is the product manager at CrunchMetrics. He has more than 7 years of experience in domains such as Energy and Utilities, QSR, Healthcare, CPG, Real Estate, Banking and Insurance and E-Commerce. He has worked on setting the strategy, developing the feature propositions, marketing the product and handling the financial metrics of the product.