So, we've talked about Volume and Sentiment. Those are the two common metrics that are reported in social media monitoring platforms. Another reason that we might use these platforms is to understand the nature of the conversation. What are the topics being discussed? What's the language being used? And so, looking at the content in detail is going to be the way to do that. Word Clouds are a popular visual tool that's used to understand content. All that Word Clouds are doing is saying, let's take a sample of documents. Or let's take all the documents and let's literally count up the number of times different words are used. And the visual representation, the size of the word is going to be based on the frequency with which that word is used. So bigger words correspond to words that are use more frequently in the responses. And this is nothing new. It's the same technique that was used for kind of the visual representation of open-ended survey questions. All right, so one problem with word clouds, it's a static point in time. So it's not a visual that lends itself easily to how does the use of that word change over time? All right, so that's one thing that we might want to consider. It is based on individual words, so it's not designed for taking into account particular phrases. It doesn't take into account the adjacency of words. It doesn't take into account potential negations. The context in which the word is used does not matter at all. And so, one of the things that we can take a look at within some of the tools, that we use, specifically within Crimson Hexagon, is well, how can we try to address these limitations, and gain a little bit more understanding about the content that's being discussed. That's how can we go beyond using Word Clouds. All right, so I've mention crimson hexagon as a tool and one of the modules in this course will kind of show you a little bit more about the capabilities of this particular tool. But what I wanted to do in terms of introducing it to you now is shows you what insights are possible using the platform, right. So, this was a study done using the retailer best buyers. As an example, trying to understand what people are talking about in regards to their expectations for shopping at Best Buy. And so, we can see these insights are coming out of the machine learning tool where someone has to sit down and train the algorithm how to classify comments. This is not coming out of the fully automated tools. So there is a little bit of investment involved in setting this up but as we'll see the insights are going to be, the capabilities are much stronger. And so we can see the distribution of comments. We see some comments talking about the shopping experience, social gratification and peer support. This is a place I can go to with my friends. Some of the comments talking about the level of service that's provided. A lot of the comments talking about convenience and availability, right. Most of those, or about a third of the comments talk about convenience and availability. Almost half of those comments talk about the products that people want being in stock at a particular point in time. And so, once somebody, once you sit down and train the algorithm, this was applied to more than one and a half million comments that were contributed over the course of a year. It's unreasonable to try to code all of those comments and classify them by hand. And so that's where these algorithms come into play. All right. Drilling down a little bit further, looking at why Best Buy shoppers are going to Amazon, all right. I get almost 5,000 comments, this being done during the holiday shopping season. So, if looking at the combination of Best Buy and Amazon in terms of the queries and we'll talk about setting up these monitors. But in order for these comments to be captured, we've got to tell the platform that we're using, in this case Crimson Hexagon, what opinions, what social media comments do you want me to look for? So, when we build these queries, we're probably going to be looking for queries that mention both Best Buy and Amazon in the comments. And among those queries, what we see a lot of the comments talk about using Best Buy and then going onto Amazon for shopping purposes. So, I shopped at Best Buy to get an Amazon gift card, about 5% of comments talking about that. But almost 30% of total comments talk about showrooming activities. So from Best Buy's standpoint, big concern for them. You know, another big component here talking about Best Buy not having products in stock and using Amazon as the alternative. Well, we saw in the previous slide one of the big reasons people were shopping at Best Buy was because they could get their purchases that same day. And so, we can use these as insights to understand our customers, to understand why we're losing our customers to the competition. And so, this is a way that a lot of marketers are starting to look at social media. It's a fast tool that can be used for marketing insights. The question that we want to ask is can the insights we get from social media how can they supplement what we get out of traditional marketing research. Just to give you an example, we're going to use Crimson Hexagon to demonstrate social media listening and to extract some data that we can use for analysis purposes. But there are a number of providers that are out there. SalesForce has a marketing suite that includes a social media component. They had acquired another popular company, Radian6. Sysomos is another popular brand. Brandwatch, Cision, all different social media monitoring platforms. Many of the capabilities are similar across these platforms. You're always going to be able to look at volume. You're always going to be able to look at sentiment. You're always going to be able to do some exploration of the content. The advanced features are going to vary from platform to platform.
