In this episode of In-Ear Insights, listen to the full, un-abridged audio from CEO Katie Robbert’s talk at the Business Analytics Conference (BACon) as she walks through use-cases for predictive analytics in multiple industries, from forecasting real estate sales to customer service call center staffing. You’ll also learn what constitutes good data and the overall predictive analytics process.

Listen to the audio here:

http://traffic.libsyn.com/inearinsights/tipodcast-baconsession.mp3

Download the MP3 audio here.

Need help with your company’s data and analytics? Let us know!

Join our free Slack group for marketers interested in analytics!

Machine-Generated Transcript

What follows is an AI-generated transcript. The transcript may contain errors and is not a substitute for listening to the episode.

So today we’re talking about time series forecasting, which is really predictive analytics.

And so the general disclaimer slide, as mentioned in the opening remarks this morning, you probably seen this a bunch today.

So introductions, I came rubber, I am the CEO of Preston sites.

I’m also a proud partner and part of the pm square data science team.

I’m a noted keynote speaker and an IBM Business Partner.

So as we’re talking about predictive analytics today, one of the things I want you to keep in mind is that it is not going to tell you exactly what to do, it is meant to push you in the right direction.

So this is a really great quote, that says the goal is not to predict the goal is to change behavior to change outcomes.

And so keep that in mind, whether you’re, you know, going through this presentation, or doing predictive analytics on your own, is meant to sort of guide you into the right direction to make a more data driven decision than just guessing which we’ll get into.

So before we get into what is predictive analytics and how to use it, I want to just quickly go through the data analytics hierarchy.

So you read this from bottom to top, the bottom being the foundation of your descriptive data, sure quantitative data, what happened, the problem is that most companies are stuck at this bottom layer.

And they can’t even really tell you what happened.

They have a hard time collecting consistent data, data that tells them what’s going on.

And then even making sense out of it reporting, we call that dark data.

So a lot of companies are sitting on these piles of piles of data that they can’t make sense of, so they can’t even really figure out what happened.

Next is your diagnostic data.

So once you know what happened, you need to understand why it happened.

So these might be your customer feedback surveys, your employee feedback surveys, any sort of market research that helps you paint the picture as to why people made the decisions that they made.

And then next, once you have those two auto wrongs, you can then safely move on to predictive analytics, which is the what will happen and that’s always that most today.

Once you have those, you move on to the prescriptive, or what am I going to do about the thing that’s likely to happen? So those are your actions.

And then I did in an ideal world, where a few years away from this type of deep, deep learning, you can find all of those together, your descriptive, your diagnostic, your predictive and prescriptive, and a deep learning model will take all of those things, and then spelled the opposite say, here’s what’s going to happen.

Here’s what you which will be fantastic.

I look forward to that day will save me a lot of time.

So why should you use predictive analytics? Well, it seems pretty straightforward, but we’re going to go through it pretty quickly.

So first and foremost, guessing is bad.

You might be saying to yourself, Well, I don’t guess my make decisions.

But one thing that I want to know that I’m guessing is that if you’re not using your own da