Speaker 1: I want to start this video with a singular, bold statement. Business questions before data analytics. And let me explain what I mean by this. In my years supporting various types of business domains, product management, and finance, and marketing, and customer service, and supply chain, IT, you name it, what I found is that the business questions, the questions that need answering most in terms of driving the success of the business are more important than the data analytics techniques. And by the way, this maxim, this assertion that I'm making is universal in my experience. It doesn't matter what business domain you're working in, the business question matters the most. And that's for two particular reasons. One, it helps you define what data you need. And two, it also tells you what broad category of data analytics techniques you should probably use to explore and answer the business question. And let me explain what I mean by that because it's kind of abstract. The first general category of business questions that I've found over and over again in my time as a hands-on analytics professional boil down to this idea of analyzing the business over time. And I'm gonna give you a hypothetical example, but based on real world experience, from the field of HR, human resources. So let's say that in human resources, we have noticed that in general, we're worried about attrition. We're worried about the number of employees that are quitting the company, leaving the company and going to work for somebody else. And let's say as HR professionals, as HR managers, we've put in a number of programs. Maybe we've increased compensation, maybe we've expanded our benefits package, that sort of thing. And what we want to know is relative to the historical performance of the business in terms of attrition, have things improved? So that's one general bucket. And by the way, that's an example from HR, but there are all kinds of examples from marketing and sales and finance and customer service that fall into this general category. So as a business professional, first thing you should be doing is formulating an interesting question that you need to answer with data and then say, oh, okay, I've got the question. Does it fall into this particular bucket? And if it does, then the tools and techniques of what I call KPI analysis are going to be useful for you to answer that question. So that's one example. Another general bucket, general type of business question that I've seen over and over and over again in my time as an analytics professional is questions where the answer is some sort of number. And I will give you another hypothetical but very realistic example. This one comes from the realm of digital marketing. So let's say that you are a digital marketer and you pay for ads on Google and you pay for ads on Facebook. And what you're interested in is understanding the effect of these digital marketing, these digital ad campaigns on sales. So sales, that's a number, that's got a decimal point in it, right? That particular broad category of business questions is also very common, irrespective of whether or not you're in digital marketing or HR or finance or whatever you might be. And in that case, the tools and techniques from linear regression analysis might be applicable for that particular problem, that particular business question. Now, once again, notice that I don't necessarily have to focus on what linear regression is for digital marketing. No, no, no, linear regression is a general purpose technique for helping you answer or investigate particular types of business questions, no matter where you work, no matter what part of the business you work in. Lastly, there's another broad category of business questions that has an answer that tends to be some sort of category. Yes, no, true, false, approve, deny, bronze, silver, gold, if you're predicting the outcome of an Olympic medal game. Those types of questions are also extremely common, and I'll give you one specifically from the area of product management. So one thing that product managers are typically very interested in these days are usage and behavior patterns within a product, and I'm gonna use a software product, but it could be other types of product. A software product, the types of behaviors and usage patterns that are exhibited in the product and how those are associated with some sort of outcome. So let's say, for example, you are a software product manager and your company sells software as a service, sells software under a subscription model. One of the things you might be extremely interested in is what are the behaviors, the feature usages, those sorts of things that are highly associated with a customer converting to being a subscription customer, they're paying you every month. Not only that, but that they tend to pay for long periods of time, what is often referred to as being sticky in the industry. So that is a prime example of using techniques from the area of machine learning, and that's another broad classification of business problems that can be answered with machine learning. This yes, no, true, false, approve, deny, sticky, not sticky, bronze, silver, gold, if you're predicting the Olympics. Once again, but notice that I picked a particular subject area, right? I picked product management. However, the general class of business question is irrespective of whether you work in product management or HR or sales or marketing or what have you. So these three general buckets of analytics techniques, notice that they are secondary. They happen after you decide what business question you're interested in answering. So if you're a business professional and you are interested in having more impact at work with data, and you're beginning to go along this data analytics journey, don't focus so much on the analytical techniques per se. The first thing that you should be asking yourself is, what are the general categories of questions that I would really, really like to answer as a business professional? Because then that allows you to do a choose your own adventure approach. So if you sit back and you think, okay, well, Dave told me that I should think about business questions in these categories, so I'm gonna do that. And all the questions I can think of fall into this analyzing the business over time category. Cool, great. Now you can choose your own adventure and say, look, I'm gonna go study the tools and techniques of KPI analysis, and that's all I need. And that's great, that's totally fine. It doesn't mean you need to learn linear regression or machine learning. If all of your business questions fall in that bucket, you know what you need, and you can just go learn those things and go forth and deliver awesome business ROI. And if you find that, well, my questions span all three categories, great, then you learn all three. And not surprisingly, based on my personal experiences in these categorizations, that's the kind of thing that I focus my content on. I focus my content on KPI analysis, I focus my content on linear regression analysis, and I focus my content on machine learning. Because in general, over time, I've found that those three general techniques, those three categories of techniques answer the vast majority of business questions most of the time. There are other techniques that are useful, to be sure, don't get me wrong. However, if you're a business professional, first and foremost, start with the business question and then map it to the data analytics techniques that you need to learn. And by the way, just so that we're crystal clear on this, everything that I just told you about, all those three buckets, KPI analysis and linear regression and machine learning, are accessible to any professional who would like to learn them. That's what my content's all about on this channel, not surprisingly. So if you're interested in seeing more applications of these three buckets to particular business domains, go ahead and click up here. I've got some videos on my channel, for example, that talk about HR analytics and talk about marketing analytics and map to these three large buckets. Also, if you're interested in learning more about some of these techniques, I'll put up a couple videos here and here and you can check them out to learn more about these various types of techniques. Once again, don't focus in the beginning on the data analytical techniques. Instead, think about the business questions that you would like to answer and then say, okay, my questions map to these three buckets and then that tells you what data analytics techniques you need to learn. All right, that's enough of this talking head video. Hopefully you found it useful. If so, would you mind giving me a like? That would be great. And until next time, please stay healthy and I wish you very happy data sleuthing.