All right, so let's just jump into the buzz monitor itself. And again, a lot of information is being provided to you on the initial screen. So this is the summary screen. And I'm going to walk you through from top to bottom. So up top you can see the timeline. Right now it's specified, recall that this query was run going back to 2012. It's presenting results from the beginning of 2015 through today. And so this timeline, the yellow timeline, is what's going to allow us to make changes to the date range that's specified. So if I want to include going back all the way to January 1st, 2012, I can drag that timeline back. The information is going to be processed. You can see over a little more than almost a four, five year period, we've got 40 million posts. But let's just restrict what we're looking at to a little bit more recent. So I'm going to put it back to January 1st, 2015. And so we can quickly change the timeline that we're looking at. And so if you want to be focused on whats happened very recently, it's easy enough to say let's look at the last week, the last month, the last quarter. If we want to do year over year comparisons, we could look at November 2015, compare that to let's say November 2016. All right, so that's the timeline that we can use. You'll see on our summary screen, we first get a measure of volume. And so we can see what the general level of volume looks like. So for Starbucks post, it seems to hover 10 to 20,000 it seems. And then there are a couple of days that really stand out during this time period. And so on this day, looks like we have almost 130,000 posts. Now, if our baseline is 10 to 20,000 that's a significant spike compared to what we usually see. We also see a spike a little bit further back in March 2015. Now from the summary screen, we don't have the ability to dig in to see what's going on. But those spikes, those deviations from the base line, that's something we'd want to probe a little bit further. As far as where the data's coming from, 94% of the post's coming from Twitter. A small fraction coming from forums, blogs, Facebook and comments accounting for a very small fraction. All right. As far as sentiment over this entire period, so the average sentiment over the entire period of beginning of 2015 through the present, 30% positive, 17% negative, 53% neutral. So a lot of comments being lumped under neutral, but almost two to one positive sentiment expressed for Starbucks. And then next to the overall average, we see the timeline, so how has this sentiment changed over time? And one thing that we might want to look at is, do the deviations that show up in our volume plot, line up with deviations that show up in our sentiment? So those high spikes for volume, what kind of comments were they? Were they positive comments, or were they negative comments? And we can see, looks like this peak in the volume of negative comments. We see a dip in positive sentiment, looks to be around the time of those 128,000 posts. So that high volume of conversation seems to have been negative chatter. Just moving down the summary page, you can see one of the new features that Crimson Hexagon has added is a motion analysis. And throughout all of the use of the Crimson Hexagon platform, you're going to see these information bubbles floating around. And if we click on them, that will give us a little bit more information about what that particular metric is reporting. And so, in this case, it's telling us six specific categories of emotions are being picked up. And for those six categories, about 25% of the posts relate to that. The other 75% of the posts relating to neutral. The other piece that I want to point out from this summary screen, again something that we're going to see throughout as we walk through the platform, are these down arrows. So if we click on these down arrows, notice we have a lot of options as far as exporting. We can export as an Excel chart, or I'm sorry, as an Excel worksheet. And that's going to contain the raw data. So in this case, if I export the volume, that's going to give me for each day in my time period, what is the volume of social media activity? How many posts? If we're interested, we could also just export in terms of the graphics. So if we want just the image, we have a number of options there. Same idea if we're looking at sentiment. We can export Excel or the image. So most of what we're going to do to pull data out of Crimson Hexagon, we're going to download the Excel files, and then do some of our advanced analysis using those Excel files. All right, so that's our summary tab. And on the left you see all of the tools that are designed for the buzz monitor. And so let's just take a glance through each of these. So if we go down into the sentiment tab, notice the way that this layout is constructed. So we have a choice of viewing the total volume of posts, or in terms of proportion. Now this is looking at sentiment. And so right now we're using stacked bar charts to display the sentiment. If I'm primarily interested, not in how many posts are positive, neutral, or negative, I might click on proportion. And that's going to change the perspective. So instead of looking at in terms of the number of posts, we're going to look at what fraction of posts on a given day. So that's one way that we can switch the view. Another way that we can switch the view, is in terms of the level of aggregation. Now I'm a big proponent of using the most granular data that's available to you. And so in this case, the most granular data that we have available to us is daily, all right. And so we can click on daily, now rather than lumping posts together that let's say occurred within the same week, or the same month, we get a much more nuanced perspective, right. And, of course, you've got options for do you want to look at it as a line chart, or an area chart, or a bar chart, all right. So again, we have our download option. Do you want to download the image or the Excel file? This is one of the cases where we probably want the Excel file. All right, now you can see, if we look at that day that had the very high level of volume and click on it, we're able to pull up a sample of the posts from that particular day. All right, so in this case, it was November 10th. These are the basic positive posts because I clicked on the green portion of the bar. If I clicked on, let's say the red portion of the same day, I would get the negative posts corresponding to that particular day. All right so these are the negative posts. Now this is giving us the post list. We can take a sample of what the posts are, we could also look at just the word cloud. Something we'll take a look at in more detail in a little bit. But this giving us just the general reflection of what are the most commonly used words on November 9th that are in negative posts. We see the words red, cup, Christmas, complain, Christian. And so this likely having to do with the cups that Starbucks began using for the holiday season, that caused a little bit of an uproar on social media, and was covered by mainstream news. And so, from the sentiment screen we can dive a little bit deeper by drilling down, and looking at the positive comments, looking at the neutral comments, looking at the negative comments. Now, if we wanted to look at the proportion focusing on that particular day, we'll see how that share changed. Before we do that, let's take a look at what the other day where we see this spike was. And we'll go over to the word cloud to do that as well. So high volume, what were people talking about? Looks like the #racetogether, that's Starbucks getting involved in, kind of more of a social movement, trying to encourage conversation. So including that hashtag on the beverages that it was serving. So we see two days that have spikes in volume. Well, the stack bar charts, one of the down sides is that it might be a little bit more difficult to see how the share of sentiment shifted during those days. So if we click on the proportion, that's going to make it a little bit easier for us to visualize. And so when we look at the proportion, we see in terms of the changes, look a little bit different. So the majority of the comments around the red cup, there was a spike in negative sentiment. Where as the spike that we saw in terms of the #racetogether, seems to have been a spike in positive sentiment. So we can see how these different events have affected both positive and negative sentiment. All right, as I'd mentioned the emotion tab, new addition to Crimson Hexagon. Picking up on different types of emotion over time, and the same types of features. So we can view in terms of proportion daily, we can download this data. And the majority of these comments coming in as neutral. So that's looking at sentiment, one of the more popular social media metrics that's reported. Let's also drill down now into the metrics section. Now one of the nice things with sentiment is, you're really getting two metrics for the price of one. You get both the sentiment and the volume. You can, however, if you're just interested in volume, we can just pull that information from the volume tab, and just as easily download it. The other ways that Crimson Hexagon slices the data, or allows you to slice the data, is by day and time, by content source. So let's take a look at those. So, if you want to see, when does the conversation peak? Well, if we're looking at the time of day, we seem to have a lull in conversation about Starbucks in the early morning hours. Seems to hit its peak at around 10, 11 AM, and then tail off after that. If we're looking at the day of the week, what's the most popular time? Looks like Tuesday, though not by a huge difference. Looks like Saturday and Sunday we have a little bit lower amount of conversations mentioning Starbucks. Content sources allow us to see where the content is coming from. And in this case, it looks like the majority of it is Twitter. Clicking on the proportion can tell us if there are trends in content, in terms of where that content is coming from. And it does, in terms of where the data is being sourced from, does look that a smaller fraction is coming from Twitter. More fraction coming from discussion forums over time. And so we might want to look at the volume of conversation. Has the volume of conversation on Twitter tailed off? Or is just that it's picked up in the discussion forums?
