Harnessing Predictive Analytics: AI for Marketing and Growth Strategies (Full Transcript)

Speaker 1: Hey, and welcome to our brand new series in which we're going to get you up to speed on artificial intelligence for marketing and growth. Predictive analytics is a form of data mining that uses machine learning and statistical modeling to predict the future based on historical data. Applications in action today are all around us already. For example, banks have been using predictive modeling to approve or decline your credit cards and personal loans. But it's not only that. It's also used for weather forecasting, recommendation engines, spam filtering, and fraud detection. So why should marketers care? Imagine if you could not only determine whether a lead is a good fit for your product, but also which are the most promising. This will allow you to focus your team efforts on leads with the highest ROI. This will also imply growing from quantity of metrics to quality of metrics, which leads to focus more time on. A financial services provider can use thousands of data points created by your online behavior to decide which credit card to offer you and when. A fashion retailer, based on the jacket you just bought, can use your data to decide which shoes to recommend as your next purchase, based on historical behavior that other customers have had in the past. But the implications are much bigger than that. Retailers can predict demand and therefore make sure they have the right level of stock for each of their products. Every time we type a search query into Google, Facebook, or Amazon, we're feeding data into the machine, growing ever more intelligent. To leverage the potential of artificial intelligence and predictive analytics, there are four elements that organizations need to put in place. First of all, you need to ask the right questions. Which questions am I trying to ask with my predictive analytics? Which metrics am I trying to forecast? Which future behavior am I trying to predict? You need a sound hypothesis to actually test. The second one is having the right data. We've come a long way in terms of data availability. It's been said that 90% of all of the world's data has been generated in the last two years. But we still do need complete and clean data sets to arrive at plausible conclusions. It's important for you to figure out what data is available to you and whether it will be sufficient to answer your questions convincingly. Third of all, we need the right technology. Whether or not a particular software is right for the problem you're trying to solve. And finally, it's the right people. Without the right people, it's impossible to pose the right questions. Let's look at staff retention at IBM. IBM is using predictive analytics to retain its employees and come up with possible solutions to forego higher turnover. By uploading a structured data file, Watson can spot the common factors in employee dissatisfaction. This then feeds into a quality score for each employee based on their predicted likelihood of leaving IBM. This is what we call people analytics. Next, let's look at supply chain optimization at Walmart. Walmart takes data instantaneously from its system and incorporates it within its forecast to assess which products are likely to go out of stock and which have actually underperformed. Combined with behavioral data from its customers online, this provides a huge amount of data points to help Walmart prepare for increased or decreased product demand. Forecasting this allows Walmart to personalize its online presence, targeting customers with specific products based on their predicted likelihood of making a purchase. We'll go into more depth on predictive analytics in our next episode. Don't forget to hit the subscribe button and be notified when episode 2 is available.