In the last session what we discussed was using factor analysis as a dimension reduction tool, if you will. So starting with a large number of survey items, how can we identify the correlation patterns that exist among different survey items and identify underlying themes that customers exhibit in their preferences and in their responses? And once we have that smaller number of themes, we can work with that rather than working with the original number of survey items. Well, one of the things that we looked at using factor scores for were as inputs into subsequent analysis. So, we had looked at using the resulting factor scores in our regression analysis to find out which preferences that consumers held actually drive purchase intention. What we're going to look at today is another technique, where the results from factor analysis can come into play. We're going to look at the problem of customer segmentation. Why do we do it, how is it done, and how can it potentially help an organization? So to review just a little bit of kind of marketing foundations, what we're interested in we have different players in this space. We've got our customers company, our competitors, and our collaborators and what we're trying to do are three different steps. It's referred to as the STP framework. First is, we want to engage in market segmentation. Second, we want to target the market segments that we've decided upon. And third, once we gone about doing that, how do we develop products and position our products in services to appeal to that target market? So we've got a segment, we've got a target, and we've got a position our offerings. Now that's one step in creating value. Then we have the four piece or the marketing mix variables that we're able to help us try to capture that value, and ultimately what we're looking to do is customer acquisition. We want to get customers into relationships with the brand or with the service provider. We want to develop those relationships and retain those customers. And that's ultimately what's going to drive profit. So as you can see, from this schematic the STP framework of segmentation, targeting, and positioning happens relatively early on In the process. So what we're going to focus on in this session, is how do we go about doing that market segmentation? All right, and so just to give you a sense for why we might be interested in segmenting the market, there are a lot of decisions that market segmentation can help drive. So if we look at the development design of products and how we position them. Well, how I position the product, the features that I include in products or services, kind of the set of offerings that I put on the market, those are all going to be dictated by which market segment I'm trying to appeal to. If I understand that a market segment that is important to us is price sensitive, I'm not going to put in the most costly features. Because the price point that the product or service becomes available at may not resonate with those customers. On the other hand, if the segment that I'm going after is particularly affluent, they're not price sensitive, and what they value is the highest possible quality, well, that's going to dictate a different product design and a different positioning. And so, the product and service design and positioning is going to be affected by segmentation, the way that we communicate with our customers, both prospects and existing customers. So, what is the messaging that's going to most resonate with them? If we have a good understanding of what their preferences are or what's important to them, what opinions they hold, what they believe on different dimensions, then we're going to be able to tailor our communication materials and use messaging that's going to resonate with them. Another important aspect where segmentation can be informative, is when it comes to resource allocation. I have a fixed amount of resources. I've got a fixed budget when it comes to the marketing activity, that I can engage in. Well, which of the segments is going to be most responsive to the marketing activity that we're going to do? Which segment is big enough? Which segment is particularly responsive? And which one's going to ultimately be most profitable for me to go after? If I've segmented the market, I can try to answer those questions and see where are my resources best deployed. Pricing is also going to be something that is informed by your segmentation strategy. Again, if we go back to the example of some customers being price sensitive, some customers being relatively insensitive when it comes to changes in price. This may allow us, if we're going after a segment that's not particularly price sensitive, but we can get away with charging a higher price or having a costlier product. On the other hand, if we're in a market where our core customers are customers who are price sensitive, well, we might have to use price promotions more frequently than we would, if we were dealing with less price sensitive customers. So, you see a lot of marketing decisions are going to be affected by differences that exist across customers, differences in price sensitivity. Differences in what messaging is going to resonate with them. What their opinions and beliefs are. Differences in what product attributes are most important to them. Those differences, that's why we're doing segmentation. We want to put customers into groups where customers within the same group are as similar to each other as possible, and customers in two different groups are distinct from each other. And if we can do that, then we can decide, all right, well. Segment one, maybe that's a profitable segment to me, let's compare that to the customers in segment two. In terms of how many are in that segment, can I appeal to those customers better than my competition, and ultimately which of the segments makes the most sense for me to go after when it comes to expected profit? So, I've given you kind of the benefits of what segmentation allows us to do. Are there scenarios where segmentation may not be particularly effective, or it might not be something that we can implement due to business constraints? So for example, what we have identified is limited variation, limited differences across consumers on dimensions that matter. Suppose that all consumers have the same price sensitivity, have very similar price sensitivity, does it make sense for us to try to segment customers based on their price sensitivity? If consumers are preferred a particular set of attributes in products, does it make sense for me to develop products that don't have those attributes, thinking that there's ultimately going to be a buyer for those products? If customers are different from each other other, then we've got a basis performing different segments, potentially developing different products, potentially using different marketing mix to go after those different segments. But if customers are too homogeneous, it may not make sense for us to try to divide them into different segments. The size of the overall market is another factor that's going to come into play. Think of each time that I form a market segment, I'm going to have to develop a marketing strategy, if I'm going to go after each of those segments. Well, if the market isn't big enough to sustain multiple segments, it may not make sense from a business standpoint to develop those different marketing plans. And so if a market is too small, we might say that it's not worth it for us to try to identify different market segments. And ultimately, what it's going to come down to are costs. There is a cost to engaging in segmentation, development of different products, development of different marketing plans, development of different offerings. It may not make sense for a business at a particular point in time to try to go after multiple segments, and they may not be in a position to devote the resources that are necessary to do that. There is a benefit to engaging in the segmentation that we can, if the resources are available, we can go after multiple segments and generate profits in multiple segments. But that's only if the costs are not a barrier for us. It may be especially for companies early on that they don't have the resources to go after multiple market segments. They may say we've done our segmentation, but really all we can do is go after one of these segments right now. So we've gotta weigh the costs associated with conducting the segmentation, and trying to target multiple segments versus the benefits associated with doing so, based on the resources that we have available today. So let me lay out the foundations for how we're going to go about implementing this STP framework. We're going to begin with segmenting the market. And as you can see, one of the first things that we have to do is identify what are the bases for forming those segments. And once we've identified how we're going to form those segments, we have to actually go about building those segments. And we'll talk about cluster analysis as an approach that enables us to do that. We also have to have a mechanism for reaching customers in different segments. And typically that's where demographics or cycographics may play a role. If we think about the retail space, retailers would typically use gender and age as a quick way of segmenting their customers. Not because that was the richest basis for developing those market segments, but those were things that sales associates would be immediately able to recognize. So if you're selling electronics, or technology products such as TVs, computers, printers, and audio equipment and somebody walks into your store, I can quickly assess male versus female, I can quickly assess young versus old. And that gives me four different segments. And if I understand what each of those combinations tend to be interested in, that's something that the sales associates would be able to act upon. So, first we have to form the different segments, then we have to decide which segments we're ultimately going to target. We said here's the different segments, here's how big the different segments are. Well, how do we evaluate the attractiveness of the segment? How do we evaluate the potential profitability that we can generate from one of those segments? So, we have to have some criteria in place to evaluate the different segments, and then decide which segments we're ultimately going to go after. And then ultimately comes to positioning products and services. What are the products and services that we can offer that are going to appeal to the segments that we've decided to target? And what's the marketing mix that's appropriate? What's the media that's appropriate to reach those different segments? So we've got the segmentation targeting, and positioning steps that are going to be involved in bringing the product or service to market. So let me give you a couple of examples, around common basis for engaging in segmentation. So this is a quick illustration of Demographics Segmentation, that was done by Experian. Looking at Daily Deal Shoppers, the break down by age and gender. So, in the purple we have the breakdown of Daily Deal Shoppers, and then in blue we have the breakdown of the overall online population. If we take a look at the breakdown by gender you see that Daily Deal Shoppers are slightly skew female compared to male. Whereas, and it seems to mirror to some extent the online population. But if we look at kind of the magnitude of the differences, we see that a larger fraction of females tend to be the Daily Deal Shoppers compared to male. Which is the opposite of what we get with the online population. If we look at the break down in terms of age Daily Deal Shoppers tending to be older compared to younger, whereas if we look at the online population, we don't see as clear a divide between them. And so that might tell us, that some of these segments in this case, we may find better luck going after say, the 25 to 34 year old. We may focus more on females compared to male, based on the differences between the representation of the group on the online population, and the representation of the group in the population of Daily Deal Shoppers. So females tending to be a little bit more likely than their online population, would suggest in terms of being Daily Deal Shoppers. Again, if we look at the break down based on age, older shoppers seeming to be more interested in Daily Deals than what would be represented by the size of the online population that they make up. So, Demographics are one way that we can split up the market and engage in segmentation. Another way that's popular for engaging in segmentation, is based on Geography. This was a map that I was able to pull, looking at in this case, it maps out the likelihood of going to the gym at least twice a week to exercise. And we can see some regions that are well above average such as the Bay area, Seattle, parts of Colorado and Wyoming. Looks like Washington DC, a little bit of New York and Boston being well above average. We also can see the regions looks like in the rust color that are well below average, and then somewhere in the middle the yellow and green, which looks to correspond to the bulk of the country. Now, a map like this could be informative, if I'm opening up multiple locations of a gym. Where does it make sense for me to open those gyms? Well, it probably makes sense for me to look at the places where people are actually going be interested in going to the gym. So, that might be Dallas. It might be DC. It might be Denver. It might be the Bay Area. All right, that tells me that there are consumers there who are interested in the service that they were offering. What that doesn't tell me is that how much competition do I already have there, what the cost of doing this business in those areas? But I at least have understanding of where are the consumers who are interested. We might have products that their demand varies considerably based on climate. So if I'm selling snowshoes, I'm not going to try to do that in the South. That might be something that resonates more with consumers in Maine, in Wisconsin, in Minnesota. Whereas if I'm selling surfboards, maybe paddle boards, that require bodies of water, well, that's going to dictate where I want to put my efforts. So we've looked at demographics briefly. We've looked at geography. Very common basis of engaging in segmentation is to use psycho graphics. If we can collect information about the activities that somebody's engaged in, in terms of what are their interests in. What are their opinions? And often, this information that we gather through surveys. We can start to paint a very rich picture of this consumer. Think back to the survey that we looked at of the auto manufacturer in the previous session. Where we had questions about people's optimism of the future, their financial freedom. The questions relating to patriotism. We start to put together a behavior profile, in terms of what opinion does this person have? What preferences does this person have? And we might also incorporate questions about their activities and what their interest are. And we can paint a very rich picture of who this consumer is, and that's the idea behind building that psycho graphic profile.