Join Barton Poulson for an in-depth discussion in this video Explore basic data tasks, part of The Data Science of Retail, Sales, and Commerce.
- [Instructor] Where you can find value in your business? And how can data science help you uncover it? Well, there's a few things you're going to want to look for when you're analyzing your data to see where the untapped potential lies. You want to look at things like return on investment or ROI. The idea is how much are you getting for what you're putting in? A more specific kind is cost per acquisition. How much do you have to spend in terms of time or money or other resources to get a new customer or to get a purchase from that customer? A common metric is sales growth.
How much are you getting right now, this quarter or this year, compared to last quarter or last year? Or customer retention rate, the idea that you are keeping your customers, they're satisfied, and they keep doing business with you. Now, these are just four possible metrics out of a very wide selection, but these give you an idea of some of the kinds of data you can be looking at when you're looking for value in your own operational information. And I'd like to mention a few other things in terms of specific tasks that you probably want to do when you're exploring your data.
They fall into the categories of web analytics, A/B testing, customer data, and predictive analytics. I'll say a little more about each of those. First with web analytics. This is information you get from your website about the people who visited, and there are a couple of very general categories. Number one is visits. You can record the number of visits. How many people came to your webpages? And then separate from that is page views. Somebody may have come and seen just your homepage or they may have come and looked at 20 or 30 pages.
And then visit duration. Some people leave very quickly, some people are there for a long time. And when you use your analytics program like Google Analytics, it will give you each of these pieces of information that help you focus on what's most effective and paying off the best. In addition to information about visits, you can also get information about conversions, where you actually get somebody to do something, like click to sign up for a newsletter or take out a subscription or hopefully purchase an item or a service. You can look at total conversions or you can break it down into subcategories, like assisted social conversions where somebody first discovers your product or your company, say, for instance, on social media like Twitter or Facebook.
They click that link, it brings it through, and you get a conversion there, they make a purchase. But it gives some of the credit for that one to the source, social media where they came from. And then there's conversion value. How much is each conversion worth in dollars or other payoff to your company? You can get this information from your web analytics program. And so that's your first step when you're looking at data for your business. Another important consideration is something called A/B testing.
This is where you create two versions of some feature on your website. Maybe you give people different offers and maybe you arrange things differently. Maybe you connect it with different options. But whatever you do, you create two versions of a feature on a website. And when people come to your website, you randomize which feature is shown to them. Some people see A, some people see B. And then what you do is you compare the performance of those two. And what it does, this is an iterative process that allows you to refine your website to find what's most effective in getting conversions ideally, and that allows you to refine your own business, your offerings, and your model.
Next, you have access to a lot of customer data. Hopefully every business is using some kind of CRM or customer relationship management software, like Salesforce or HubSpot or Insightly. There's a huge range of choices, but the idea is you need to have data on your customers and you need to have a centralized location. CRM software lets you do that. It lets you organize leads, potential sales. It lets you track interactions and conversions.
It lets you have a central place for gathering information about people. A lot of these software programs will actually tag these people and gather their information from various social websites to give you a more complete picture of them. And so these are some of the most important concepts and some of the most important metrics that you can look at when you're evaluating your company data for improved performance in retail sales and marketing.
- Data fundamentals
- Retail, sales, and marketing data
- Analyzing consumer sentiment
- Exploring market segmentation
- Evaluating leads and customers
- Getting recommendations from engines
- Pricing dynamically
- Predicting trends