I'm a marketer, writing a post about predictive analytics tools, but I'm a little confused.<p>Here are some of my questions that I'd really love to read your answers to:<p>1. What does a predictive analytics tech stack look like? Which tools do you use for which purpose? Examples would be phenomenal!
2. Are there alternatives to the above tools for SMBs?
3.How much data do you need to have until your predictive analytics gets smart enough to predict things accurately?
4. What are the typical "triggers?"
5. What are some practical, lesser known ways to use predictive analytics in marketing/CRO? What can you do with these predictions?
6. Anyone think predictive analytics is BS? If so, why?<p>Sorry for all the questions. I'd really appreciate any and all insight or points in the right direction! =)
I run aihello.com that uses predictive analysis to predict which of your inventory is going to sell where and how much. We use this stock up appropriately in recommended geographic area. (for example 30% more organic products in the east coast vs west coast etc)<p>We use sklearn, facebook prophet and apache spark ML<p>Here is a screenshot of the dashboard of one of our customer : <a href="https://screenshots.firefox.com/hO1TIVsQWoNolqLj/www.aihello.com" rel="nofollow">https://screenshots.firefox.com/hO1TIVsQWoNolqLj/www.aihello...</a><p>>3.How much data do you need to have until your predictive analytics gets smart enough to predict things accurately?<p>Depends. We have roughly 4.5 million data points for each month and the results are practically usable (Customer has no issues with the error margin in predictions)<p>>4. What are the typical "triggers?"<p>Don't understand this.<p>> Anyone think predictive analytics is BS? If so, why?<p>This is like asking if anyone thinks trigonometry is bullshit.