Scaled Inference Blog The Scoop on Machine Learning for Optimization

It's a very exciting time to be a growth marketer. Because, for the first time in a very long time there's a powerful new tool in our arsenal--machine learning. You've probably heard a lot about AI, Machine Learning and maybe even Reinforcement Learning. But, you may not have had much support for understanding how this technology can be used to benefit your growth initiatives. If you are not an engineer, data scientist or on the product side you may have believed this new technology was out of reach.If you have a digital property you need to be using Machine Learning. All of the unicorn companies are using it, Growth Hackers are discussing the secret to their rapid experimentation and the driving factor for growth (guess what their secret is)...Machine Learning.The traditional months of A/B Testing are still of value (ie. thoughtful variant design), but for exponential growth Machine Learning is key. Think of it as A/B Testing on steroids where 10x improvements are a common result. After all, only 1 in 8 A/B tests have been found to drive statistical change- so let's change the status quo and amp up our experimentation culture by utilizing proven technology.At Scaled Inference we like to use an ice cream analogy to explain how Machine Learning for Optimization works. (fun fact, if you meet us at an event we may have ice cream in our booth).So, let's pretend that you are the teacher of a first-grade classroom. You want to bring the class a sweet treat- ice cream it is! You ask the class "How many of you like chocolate?" 60 % raise their hands. You ask "How many of you like vanilla?" a smaller percentage of hands go up. In an A/B Testing framework this real-world example would lead to everyone in the room being served chocolate. Chocolate caters to the majority making it the "winning" variant.But now there are disappointed children in the room and no teacher wants a revolt on their hands. So what to do? What if this real-world example was based on a Machine Learning approach? We would use context to better serve the audience.In the first graders' classroom perhaps the context and data available are the school health records. Perhaps some of those who declined to raise their hands for ice cream preference have a hidden decision factor. So, you look at the records and low and behold you learn that 3 of your 20 students have a dairy allergy. Now you could offer a lactose-free (new variant) option. This means that in our classroom example we have gone from serving approximately 60% of the population to serving all of it (now that's sweet!).This was just a little sample, a taste for Machine Learning (AI).Want to get the whole scoop? Watch our Free Webinar and discover how you can use AI for personalizing customer interactions in your organization.