Zia can understand business metrics, analyze them, and predict the possible outcome that can help in decision-making and preparing your business for the future. To substantiate the prediction results, Zia shows the analytics that were taken into consideration to predict an outcome.
Zia Analytics
Zia Prediction is designed in such a way that it helps sales reps improve the quality of their day-to-day activities. It helps managers take in the prediction insights and help improve their decision-making, including which deals to focus on, understanding where the sales reps need to improve, and more. With the existing prediction analytics, users can understand Zia's prediction capabilities and its broken down for users to better consume these analytics.
Overall accuracy of the predictions and its quality:
Zia will give itself a score on the strength of its prediction capacity. It will take into account the prediction score it gave, the number of deals closed based on the prediction, and the number of deals lost. This way, Zia will learn to correct its own trajectory to better study the data and provide better predictions in the future.
For example, 50% accuracy is considered acceptable, while more than 50% is excellent.
For example, in the image below we are predicting the type of health insurance that is likely to be purchased by a customer. Based on the analytics, Zia shows that 2465 active predictions are currently available and the accuracy is 38%.
The number of records that uptrend or downtrend:
Based on the recent activity a record's progress towards higher or lesser probability is calculated and displayed as an uptrend or a downtrend. For example, in the image above, five deals uptrend and two deals showcase a downward trend for deals which are about to be lost as predicted by Zia.
Time-based graphical representation of prediction accuracy display:
You can see the number of accurate predictions made last week, month, quarter etc. You can also view the number of successful and failed predictions, calculated by evaluating the actual field value vs. the predicted field value at the end state. For example, the number of deals that were predicted as closed won vs. the actual number of deals that were closed won. As per the image below, 32 deals were accurately predicted.
Failure prediction - Users Who Got Most Prediction As Failure:
In this, Zia will provide a report on all of the total predictions it's given for each user in the CRM system, the number of predictions that failed and the number of active predictions in a tabular format. This will help admin-level users understand how users are approaching the deals and aid in better decision-making. For example, if Zia has predicted a win for a particular deal and it ended up becoming a lost deal, managers can understand what went wrong with this analytics and even go to the extent of deciding whether to rely on this prediction or not.
Closed Lost Prediction Deals Current Stage:
Whenever sales reps are focusing on deals and it gets to a point where the deal is going to be lost, it would have been better if they got an idea beforehand so they can focus on those deals better. Zia will predict deals that are likely to be closed lost based on its current stage and this depiction can help sales reps focus on what can be done better at the early stages instead of when it's usually too late.
Win/Loss analysis:
Zia Prediction will provide the win/loss analysis, taking into account the numbers that went into the prediction, and how efficiently the system helped in predicting a win or a loss apart from that, it also lists the most influential factors for arriving at this analysis. For example, take the case of a Health insurance company, their win/loss analysis looks something like this. Here you can get the overall health insurance policies won by reps and those that are lost, the number that was predicted and the success and failure rate of the predictions along with a graphical representation of the same data.
Failed prediction segmentation:
Zia will segment the previously failed predictions, and validate them and this information can be used to understand the common characteristics that different segments have to arrive at a pattern for failure. It will group the failed predictions as segments based on different fields. You can filter to see the unique fields and they will be highlighted like this:
This can help you assimilate a reason for failure if any and take action accordingly.
Contributing factor:
Every record will have a Zia prediction contributing factor to help users understand data at a record level. But to understand the overall sentiment of the organisation, Zia prediction analytics will now give you an overall idea of the top positive and top negative contributing factors that have influenced the winning deals and the lost deals respectively.
This again will help users understand where they are and perform course corrections if needed.
Delayed deals:
This will help users understand which deals were delayed beyond their prediction dates. Zia will provide a report on the predicted date and the actual date of conversion along with the Deviation in terms of days.
Points to remember
- If there is no sufficient data, Zia will not be able to compute the analytics. Hence, there won't be any data to display.
- If the number of records in active prediction are shown as 0 that implies the records may not have undergone any recent update or they have been newly added to the CRM system. You can wait for a couple of days or weeks and revisit the Analytics to see the uptrend or downtrend.
To view prediction analytics
- Go to Setup > Zia > Zia Prediction.
- Click the Analytics tab.
Note:
Only those users who have the permission to manage Zia configurations can view the Prediction Analytics.