Speaker 1: Hello, everyone. This is one of the marketing engineering sessions. We are going to learn conjoint analysis today. And the conjoint analysis is related to new product development. So a lot of marketers use conjoint analysis when they develop a new product, okay? So let me get a file to show you an example. So I have OfficeDar example again. So I have data set, intermediate data set, and you can see conjoined study design here. So it has four different attributes you need to consider when you develop a new product, right? And there are certain levels in each attribute. For example, location has three levels, so close location within two miles, intermediate two to five, and far location is five to 10. In office supplies, you have three levels, very large assortment, large assortment, and limited assortment. And furniture too, yes or no. And computer, three levels, no computer, software only, and both software and computers. So you can make a lot of combinations using this table, right? So below, you can see bundles. So these are the combinations, right, out of this conjoined study design table. So altogether, possible combinations are, bundles are, same thing, 16, okay? You have 16 different bundles here, okay? Combining four different attributes and, you know, two or three different levels of the attributes, right? And these are the ratings from respondents, about 16 different bundles, okay? So each respondent rated each bundle, right? For example, number one person, 90, 50, 50, 80, 85, 40, something like that, right? So please note that they rated the whole bundles, not particular attributes, no, okay? They rated the whole bundles, whole products, okay? Right? So we can go ahead for analysis and add-ins, marketing, engineering, and it's a conjoint here. With the data set, you already went through create study design and create data collection. So you can go to the third step, okay? Estimate preference part words. Click that. All right, so pop-up screen here. So I don't have to touch anything here. So go, click next. Highlighting, yes. So this is the result, respondents preference part words, okay? So these are the part words of each level, okay, from each attributes, okay? Somehow the software calculated the part words of the levels of attributes based on their ratings of the whole bundles, okay? So this is the beauty of conjoint analysis, right? So the consumers rate the whole bundles, but you know, you can find out which level of each attribute is more important for the consumers, right? So this is the meaning of part words, okay? All right, so from it, you can go to next step. This is a little bit more complicated, right, than other methods, right? So you can create analysis template, right? Existing product profiles, yes, you need to click new product profiles too, right? So with existing, you will consider new product, right, okay, next. Also highlighting, right? So you will see a pop-up screen here. So existing brand, so you can, I'm going to fabricate data here. So A is the existing product, and it's far, and larger assortment, no furniture, and no computer, okay? So I can add it to list, okay? Another B has, okay, intermediate and large assortment, and less furniture, and both, that's two, right? So okay, so yeah, to make it simple, I have only two existing products here, next, and then new, okay? I type new here, location close, I'm going to have more ideal levels here, large assortment, very large assortment, and less furniture, and both, software and computers, add to list, okay. All right, so these are the profiles of two existing products, and new products, okay? So I can go one, two, the final analysis, okay, run analysis, right? So existing product only know with new product profiles, right, so you need to check this. First choice rule, same thing, you know, we learned this from positioning technique, right? First choice rule means we know text at all, and share of preference, they can share of, you know, different, share of the preferences, right? But let me check first choice rule, right? Okay, also highlighting, right? So what we get is this, right? Predict the market share between two existing products, B had 95, and A had 5, okay? This can be calculated from the part works, right? But when you add your new product, right, things will be changed, oh, yeah, okay? Because I chose all the ideal levels, okay, of the attributes, so it can get it all, 100%, right? So it's not real thing, but, you know, I showed you how to do conjoint analysis with this example, right? Thank you.