Introduction to Business Analytics (Updated Edition)

so let's talk about business analytics an extremely popular topic right now but before we dive in let me mention if you want a cheat sheet that covers a lot of things that we're going to discuss here you can download it at codybaldwin.com all right so first in the video we're going to cover what is analytics some of the basics then we'll talk about the different types of analytics the life cycle which is kind of like the scientific method some of the more popular tools and the careers in analytics okay so what is analytics at a

basic level in analytics our goal is to turn data and sometimes lots of it into meaningful business insights something that can help us grow and improve our business that's really all it is however in reality it's never quite this simple so in analytics often it looks something like this we get a lot of suspect data that requires plenty of cleaning and scrubbing before it can even prove useful and after we do our analysis not all the output is helpful some of it's just noise as we call it it can be a real challenge now at

this point i should acknowledge too that the minds of many analytics is a lot about showing beautiful charts graphs and dashboards and these visual aspects are part of it but analytics is a lot more than that so here are a few specific examples of business analytics let's say you work at a credit card company you could analyze data about your customers to determine who might subscribe to a credit card offer that would allow you to spend more time and energy on targeting those specific customers you could work at a tech company and you could review

data about your employees to understand why they leave and then take action minimize turnover because turnover can get really expensive or you could work at a regional bank and you could review data about your loans in order to predict what customers are most likely to default so the options are endless these are just a few examples of how you could use data in your business to turn out or churn insights now before we go further i should also mention that there's a lot of different terms you're going to hear around analytics like business intelligence decision

science data science data mining there's a bunch of them just know that there are some differences but for the most part they all support the same goal turning data into useful insight so don't get too caught up on the differences between all the terms now let's talk about the different types of analytics there's really three key types the first is descriptive analytics and that deals with what happened in the past we're reviewing the past predictive analytics is what might happen in the future and prescriptive is based on that prediction of the future what should we

do okay and so it gets more value added as you go down this list and here's some examples of each of these types of analytics okay so think about descriptive looking in the past we could ask questions like what were our sales what was our market share and what product was most popular descriptive is looking at the past that's still an important part of analytics then there's predictive and prescriptive which is a future look what are the expected sales what is the expected market share what product should we market to our customers it's all forward-looking

okay so those are our three types of analytics now let's talk about the analytics life cycle and you can think about this as like the scientific method for analytics it's a tried and true way of doing things and the most common life cycle is what's called the crisp dm the cross industry standard process for data mining and there's really five pieces to it you start with a business understanding understanding your business problem then you go to data understanding then data preparation getting the data ready doing modeling which we'll talk about and then evaluating your model

and deploying it actually using it in the real world now what's important here is you and the reason why we use this life cycle is because if you rush into analytics you might just crash and burn you don't want to spend a bunch of money on analytics without having a business goal that you're trying to achieve okay so let's talk about these five aspects of the life cycle or phases in a little more detail so let's start with business understanding we should always start with business problems we aren't just doing analytics for fun we need

to have some business goal in mind some examples might be to optimize pricing to boost revenue or to segment customers to tailor product offers to them or pinpoint bottlenecks and failure points in our supply chain these are some business problems that we might be trying to solve using analytics now let's talk about the next phase data understanding and a lot of what we're doing here is looking at what data we have and what data we need and trying to cover some of those gaps and then we get the data we might ask questions like what's

the availability of the data the quality of it the granularity how deep or detailed does it go what's the frequency of it how often does it get updated and so on now as we're trying to understand and explore our data oftentimes we use a sandbox which is a safe space to explore our data so we don't mess up what's called production where all the live data is we don't want to accidentally delete something but we're trying to understand our data then there's data preparation the next phase and oftentimes this can be the most time consuming

piece of this it can take a lot of effort to clean and scrub your data to get it ready for further modeling and analysis so here's an example maybe you get some data like this don't be surprised if you see something like this now you have two customers one of them has a city that's missing the states are in different formats and also the date of birth and so what you might have to do is do cleaning to get this in a good place in order to do the next step to do your modeling now

what you also might want to do at this point is go back to the technology teams and say let's make sure we can restrict the inputs that we're getting so we need to do analysis we actually get good quality data now after data preparation we're going to go to the modeling phase now a common question i get is what actually is a model here's a definition of it a simplified description of a system or process to assist calculations and predictions that's a mouthful here's the way i would interpret that a model is something that mimics

the real world it's our version of it maybe when you were a kid you built a lego model that looked like the white house or the pyramids in egypt that was a model of the real thing as we try to make predictions we try to build a model and we try to see if we can get a good but also simple model to help us make our predictions here's an example of a model a model that predicts the likelihood that a car insurance customer will get into an accident in the next year a model that

makes that prediction to mimic what might happen in the real world and so in this modeling phase we're going to do a few things we're going to do exploratory analysis on the data we're going to do variable selection to figure out what variables should be included in our model and then we're also going to select the model and then fine tune it now common modeling tools we're going to use here are python or r those are probably the two most common modeling tools now after we've defined our business question and then understood and prepared the

data built our model then we're going to evaluate and deploy the model so we ask questions like how effective is this model is it working well are the predictions fairly accurate and if so are we prepared to launch it so once we build our model we can't just walk away we actually have to start using it now let's talk about some popular analytics tools certainly this isn't all of them but this is a few of the most popular ones okay so we might use in business analytics microsoft excel to help us explore and analyze smaller

data sets we might use tableau desktop to help us visualize our data using dashboards we could use the python programming language to help us build these models to make predictions that we just talked about and then we could use sql to allow us to communicate and interact with databases so it's not uncommon for job postings and business analytics to cover a lot of these tools but i will mention there's a couple other tools for visualization and for model building that aren't listed here and i'll switch to those two so you could also use in addition

or in place of tableau desktop microsoft power bi and instead of python you might use the r programming language so now let's talk about different careers in analytics for the most part in business analytics you sort of ride between a few different disciplines obviously you're probably going to be closer to business but you're also going to have skills with technology and with math now to give you an example of some other jobs kind of at the end of these spectrums in business you might have a sales rep and technology you might have a database administrator

and a math you might have a statistician okay so they're going to be at the end of these spectrums in business analytics we're probably going to ride right in the middle maybe closer to business perhaps now in analytics you could sort of fit into two different categories so it could be that the job is a business role but is supplemented with analytics and there's going to be a lot of jobs like this like finance professionals are going to need more analytical skills or it could be a role that's just analytics that's sort of its focus

now here are some of the common job titles with those analytics roles could be a business analyst a business intelligence analyst an analytics manager or a data analyst and possibly a data scientist and i put an asterisk at the end of that because sometimes those data scientist roles require more of a technical or more of a math background it doesn't mean those are all that way but these are just some of the possible roles or job titles you would see for those analytics positions within that category now with most of these job postings they're going

to mention software so you've got to know the key tools and we talked about several in the last section the video here but an easy way to develop those skills and to strengthen your resume is to select a tool download a free version or a free trial many of those tools have that get a pizza and spend a weekend to learn it so what you can do is certainly you wouldn't say you're an expert in these tools but you have experience with them that's an easy way to start strengthening your resume preparing for some of

those positions all right thanks for watching just as a reminder if you want a cheat sheet that's summarized a lot of things that we just talked about in the video you can download it here