Freshworks Analytics offers AI-powered predictive capabilities in the form of Forecasting for Freshdesk. With Forecasting, you can now see projections for how your operational metrics would turn out for your current month and the next. The forecasting reports are backed by robust AI algorithms, which are built by taking into account several factors such as historical data, trend analysis, seasonality, etc. The algorithm is a state-of-art Machine Learning algorithm that continuously learns and improves over time.
Forecasting enables Admins on Freshdesk to predict their incoming ticket volume, first response time, and resolution time metrics based on historical data, Trend Analysis, seasonality, etc. It helps customer support admins make business decisions and achieve operational excellence.
Highlights about Forecasting:
- You can select the time range of data (1 Day, 1 Week, 1 Month, 3 Months, 6 Months, 12 Months, 15 Months, 3 Years), frequency (Daily, Weekly, Monthly), and also the confidence interval. (Confidence Interval gives the range in which prediction will lie for selected value (probability). For example: Selecting a 95% Confidence Interval implies that there is a 95% probability that the forecast will be in the given range.
- Forecasting predictions are made by taking into consideration several different factors like historical data, Trend Analysis, seasonality, etc. The Algorithm is a state-of-the-art Machine Learning Algorithm that continuously learns and improves over time
- When viewing the Forecasting charts, you can select up to 5 different filters to slice and dice the data the way you prefer. The filters include Priority, Group, Source, Ticket Type, and Product. The filters will only have those values in the dropdown where there is at least one combination with predictive values. For example: If there is only one ticket with Source as Email in the last two years, then you will not see Email as one of the drop-down values under the Source filter.
- For Forecasting to be enabled for any combination of filters applied, the corresponding data should be available in the last 3 months. For example: If we select filter Priority to be High and Source to Email in the filter selection, then the forecasting engine needs at least 150 days where High Priority Email Tickets are present over the last 2 years for accurate predictions.
Accessing Forecasting Report
‘Forecasting Report’ is available as part of Curated Reports. You can access it from the Analytics homepage or using the Instant Navigator
In Curated Reports, You can find forecasting charts for three important operational metrics: Tickets Volume, Average Resolution Time, and Average First Response Time.
Note: Freshworks Analytics can predict only tickets that are assigned to any group and not directly to an agent without any group.
As of now, users will not be able to clone, schedule, or share the Forecasting Report.
The prediction is refreshed daily till the next calendar month, and model retraining happens every week so that the forecasting algorithm can improve over time.
If you're interested in trying this feature out, please sign-up for our Early Access Program (EAP) by filling out this form. We will reach out to you in the coming weeks.