Thanks to those that attended our session on Leveraging Machine Learning and Predictive Analytics in Your Forecasting Process. I’d love to hear more about your forecasting journey and challenges you have today with Forecasting! Is there anything specific you’re looking for in a forecasting system?
To access the recorded version of our session, visit this URL and enter the same credentials you used on the day of the Virtual Summit. Enter the Theater section. Then go to Breakouts, find the session name, and click 'Watch On Demand.'
We limit the number of people who work on a forecast so that we can keep things straight a little more. It would be nice to be able to put in a series of assumptions and then just have it run.
I am definitely going to watch the recorded version even though I did attend the live version - there were so many good points you made that I want to revisit! I haven’t had time yet to build out a forecast model in Prophix so I’m still using an excel model and uploading into Prophix. I’m trying to get as many ideas as I can from other users who have their forecasts already being done in Prophix!
No matter what you use to forecast today, Machine Learning and Statistical methods are extra tools in your toolbox. Happy to hear that you enjoyed the session and I would love to hear more about how you currently execute on your forecast.
I’d love to hear more about what techniques you’re doing today and how we can help improve your process in Prophix - maybe you can post again after watching the session.
Great to hear you’re using predictive modeling in your forecast. Maybe you can elaborate a little more about how the predictive piece of your forecast helps with your overall process?
We currently do not use Prophix for that. Our predictive analysis uses historic customer spending trends using machine learning tools. We analyze each MCC ( Merchant Category Code) and use that to predict future volumes more accurately.
Yep. We aren’t at the machine learning place in our org yet (hoping to tackle RPA’s soon, then evolve from there), but I’ve set up a bunch of our forecasting models to use statistics and regression models based on operational drivers. It’s actually pretty cool when one works out so well that, you’re like, whoa. Our in-month forecasts have become gospel, and they’re always within a couple percentage points now. Before I came onboard they were all over the place.
RPA is such an interesting area right now. Thanks for sharing a little bit about your forecasting models. One point from the session is that statistical models aren’t going away if they work well for your business case. Your on the right path if you already have a predictive model in place and you could layer on or try out Machine Learning when you’re ready.
I’m really excited about the features that were previewed in this session! We currently use a combination of judgmental and statistical methods in our forecasting. We rely almost entirely on historical data and make tweaks based on things we know like an upcoming promotion, the weather, what our competitor is doing, etc. but it is a very manual process. I think that if we could incorporate machine learning it would be possible for us to find a better model that we could feed different types of data into and get a better forecast. I’d love to hear what others are doing and learn about any resources that could help us improve our forecasting while we wait for these awesome features to be available in Prophix.
Really interesting breakout session. Really enjoyed digging into the various methods of forecasting options and how machine can take some pain out of the equation.
