Instructor: David Schuff, Section 003

Weekly Question #9: Complete by April 20, 2017

Leave your response as a comment on this post by the beginning of class on April 20, 2017. Remember, it only needs to be three or four sentences. For these weekly questions, I’m mainly interested in your opinions, not so much particular “facts” from the class! If you sign in using your AccessNet ID and password you won’t have to fill in the name, email and captcha fields when you leave your comment.

Name and describe a business question that you could answer using clustering. What data would you collect to perform the analysis? Don’t use an example we’ve covered in class.

0 Responses to Weekly Question #9: Complete by April 20, 2017

  • A software developer can use clustering to gain insights on the type of clients that he or she provided the software for. By performing clustering analysis, the software developer can tailor his or her product according to the type of clients using the software. For the analysis, I would collect data about age, income, occupation, and client’s company to divide the customer place into different groups.

  • Clustering can be used in real estate when determining what price range of houses are located within a zip code. The data needed for this analysis would be the cost of the house or the price is was last sold for. This analysis would be useful for homebuilding companies and would help them determine if they should build a new community in a certain area/ what range of houses would be a good idea to build.

    • I think that this is a great idea! It might also be important to keep track of the size of the house, and general features such as number of bedrooms. Construction is an industry that could definitely make more use of data.

  • A business question that can be answered using clustering “what types of shows should we recommend to viewers?” (a question Netflix would ask). The data that would be collected are: the shows’ genre, the origin of the show (which country?), and whether or not the individual watches the show with subs/dubs. Instead of collecting data on the geography, race, etc of the individual, – which is useless data – collecting data on the tastes of the viewers would be better in clustering.

  • A business question I can think of is where to place a franchise. The franchise would need to be placed in an area that would give it the highest success rate. So data to collect would be possible locations, median income for those locations, competitors, demographics, and population.

  • Clustering can be use for finding the best sold cars so the car dealers can import more of those cars. The data needed to collect would be the age of buyers, their gender, whether they are married or have children, the past history of cars sold and their regions. The car makers can also use this clustering to influence their decisions on which cars to make.

  • LinkedIn might utilize clustering for its analysis purposes. There are many types of LinkedIn users: recruiters, casual users, regular participants etc. LinkedIn might use clustering to differentiate between these types of different users. This information can then be used for targeting and strategic purposes. The data that the company would likely need to collect to perform the analysis may be: number of connections, frequency of sign on, job title etc.

  • Clustering would be very useful in marketing for different demographics. For example, Gap owns Old Navy and Banana Republic, and all three need to target different customers. Clustering people by age, income, if they’re a parent, job type, and where they live could determine this.

  • Clustering could be used to find the best locations for specific types of grocery stores, such as Whole Foods. Useful data would include population density, median income, number of other grocery stores in the area, number of farmer’s markets in the area, space available for the store, etc. This data could be used to determine if an area is suitable for a Whole Foods.

    • Lauren that’s a great idea, I feel that this would enhance the decision making process for introducing new stores to different areas. In addition I also feel clustering could be used in the operations of the Whole Foods store once it actually opens. Determining when the busy times are according to frequencies and times, and also comparing and contrasting prices for certain products along with sales and discounts and how they each affect the result of customers in the store.

  • An example of how clustering might be used would be when police try to determine where to always patrol or where to place a new police station. This would be determined by clustering the different crime statistics to determine where the most crime happens or where a certain type of crime occurs most. They could cluster based off type of crime and send the people best fit to deal and investigate those crimes to those specific locations. The data that they would need would be the zip code or location, a list of crimes committed in those areas and the number of times a particular crime occurred.

  • Clustering could prove to be helpful to retailers. Retailers such as Target offer a wide variety of products and through the use of clustering, they could determine different groups of customers and advertise accordingly to each. for instance, they could be clustered using factors such as age, type of purchase, and frequency of purchase.

  • Real estate investors can use clustering to market properties to different demographics. For example, an smart investor would cluster their prospective buyers based on each “groups” demographic (e.g. Age, income, job, relationship) in order to target them with homes and properties that are relevant to that group. A rich, 50 year-old couple would be more interested in a quiet property in Rittenhouse; whereas, a 22 year-old single college grad would be more inclined to live in Manayunk or University City area.

  • Clustering is definitely a useful analysis for a clothing store, like J. Crew or Banana, and how often customers use their loyalty/membership cards. They can be clustered by name, visits, gender, age group, etc. By using cluster analysis, we can find more information on customers by certain groups.

  • An example of how we can use clustering to answer a business question would with a grocery store loyalty programs. If the store wanted to know what type of shopper you are, they could group shoppers into 4-5 broad categories based on what they purchased. Some shoppers may only buy soda and snacks, while other will get various other types of product. This would help the store properly distribute coupons and better understand which products to discount, appealing as many customers as possible.

  • Clustering could be used for property insurers. They could gather information on property losses to see where losses are occurring, and if they are common losses, it can show an issue with the property materials or location. They could also use the amount each loss costs to see which losses cost them the most, and that can help them better quote premiums for their clients.

  • A business question that could be solved with clustering is a companies target market. The data collected would be customer demographic, purchase, and psychographic information such as age, gender, amount spent, frequency of purchases, and what types of products were purchased. The clustered data would provided information on which similar groups spent the most, spent most often, and which product they buy. This could help define who their target markets are for new marketing campaigns/strategies.

  • Clustering can be used to do research on schools located in a specific zip code or county. Variables could include price, education quality, student population, number of classes, etc. These clusters could be factors to where a new family moves and where they would want to send their kids to school. It may prove that a lot of the schools are the same, but it could also single out certain ones that perform above average. A diamond in the rough.

  • Clustering could prove useful in a retail setting. This could be used to better predict what customers will buy what items and later help them send emails targeted to different shoppers. Retailers could collect information like age, gender, zip, items bought, etc.

  • Clustering can be used by the casino business. Casino owners can gather information about demographics, buying patterns, household income, etc. to determine where to build a casino. They could also use age, credit history, and internet browser history in their clustering calculations.

  • Clustering is useful when it comes to supermaket stores. Comanies can gather customers’ buying patterns from their loyalty cards in order to understand their buying patterns and which food categories they purchase most from in order to target different customers with different coupons or promotions.

  • A bank would use clustering to understand the profile of its customer base to create targeted campaigns. The data collected would consist of the customer’s balance at the bank, the number of transactions done in the last x amount of months, the change of the account balance in the last x amount of months, the customer’s demographics, and the customer’s total balance will all U.S. banks. The bank can map age and income to the clusters to create a story of the clusters that would tell which are high potential and low potential customers.

  • Clustering can be used for businesses wanting to know where their customers live to determine where to advertise more. The business would have to gather the customers city, zip code, state, and country. This will allow the company to use clustering by using different information.

  • Clustering would be useful for an automotive retailer. The store could have prospective customers fill out a short survey providing information on how old their car is, what type, where they live, etc. This way, the store can cluster the customers by their probability to buy a new car and pair their best salespeople with those customers.

  • Using clustering is important in healthcare insurance companies when it comes to pricing insurance products. The data that actuaries usually collect include age, gender, income, job type and past health condition. Insurance companies use that to determine whether to issue policies to customers. For example, healthy customers will get cheaper premiums than unhealthy customers; Customers with steady income will have less investigation from insurance companies for the claims process.

  • Clustering could turn out to be useful to retailers. Retailers, for example, Wal-Mart offer a wide assortment of items and using clustering, they could decide distinctive gatherings of clients and publicize appropriately to each. for example, they could be bunched utilizing components, for example, age, kind of procurement, and purchase of the time.

  • There are many businesses that could use clustering to help them target specific customer groups. The most obvious examples are retailers clustering groups of shoppers based on gender, age, loyalty status and sending promos that appeal to these groups.

  • Clustering can be used in many instances, one great example would be business location. When opening up a new business or searching for the perfect location of where to place that facility, you might want to gather all of your data and put it into cluster form so that you can vividly see the areas in which your customer segments flock towards. This is a strategic way to ensure that customer traffic will flourish in the area.

  • Clustering could be used for creating an efficient delivery service. For example, if a pizza shop wanted to decide how to best allocate x amount of delivery vehicles, they could do a cluster analysis based on their historical delivery data. Specifically, they would analyze all of their previous delivery points with x amount of cluster groups. This way they can establish the most efficient way to allocate which drivers are in charge of of which delivery orders.

  • Clustering can be used to see if a specific store is suitable for a geographic area. The things that would be useful to know is population density, income, number of competing stores in the area, and space for the store. All of these would be useful in determining if a store should be placed in that location or placed in another location.

  • Clustering would be useful for recommending Youtube videos to people, based on videos that they’ve watched already. They would gather data from what you type in the search bar, along with what recommended videos you click on, along with your name, email, and age if you make an account.

  • Clustering could be used by employers to find potential candidates. An example would be having an employer on Foxnet use clustering to pick out candidates who applied for their position. They could cluster based on the qualifications they are looking for and see how many candidates come close to a certain number of qualifications. For this, they would need information on resumes.

  • With the NFL draft coming to Philly this year I figured NFL recruiters could use clustering to determine what players to draft. They could collect data such as colleges attended, positions, and player stats to help pick out top prospects for the league. For example, they could use clustering information to find college quarterbacks with the most potential to be successful in a professional career.

  • Clusters can be used to analyze inventory patterns in a company’s SKU profile. This would help management identify inventory across similar and different categories to better allocate resources when it comes to re-slotting a warehouse, determining labor schedules during different seasons, etc. Beyond the warehouse, it would help companies better understand the needs and requirements of the business and make determinations on ordering product/material from suppliers based on the data. With this information, companies can streamline procedures, improve efficiency and reduce costs when fully understanding the behavior of their products.

  • A scenario where clustering can be incredibly useful is when a mall wants to divide customers into groups to study how to organize its organization of retail space. The type of stores where they make purchases, how much a consumer paid (if possible), and the time between purchases, and other variables can be used in a cluster analysis. The result should allow the mall to identify a number of groups with fairly different choice of stores, purchasing power, and mobility. They can find a group of problem customers (Ex: those who linger long, move around a lot but do not buy much) and other useful insights.

  • Clustering can be used as a perfect analysis when determining the type of specialty stores that should or can be placed in certain areas. Specialty stores need to be placed where they are going to have the highest success rate among consumers. Owners of those stores are taking a look at clusters of data such as demographics, age, lifestyle, and income from prospective customers that populate certain areas which can all be used to determine the necessity of a specialty store placement.

  • A business question that could be answered through the use of clustering is if a company such as TicketMaster, or Stubhub wants to determine what type of individuals are purchasing tickets to certain concerts. In this scenario you could have groups by age to determine what type of crowds are being dragged in depending on the artist performing. A ticket company could then forward this information to the artist to help them further understand their audience base and what type of groups they are reaching.

  • Clustering could be used at a University such as Temple. They could use clustering to see what students are apart of what major, where they are from, age, gender or socioeconomic status.

  • a business question that can be answered by clustering is what type of customers who will leave and switch to competitors’ services next year. Clustering can help us develop a profile for this type of customer and prepare for customer attrition or come up with attractive campaigns to keep customers

  • Police could use clustering to help determine where crimes are going to happen. They can keep track of the location of the crime as well as the type of crime, time of day, etc. From here they could determine hot spots that need to be patrolled more.

  • clustering analysis can be used on shopping websites to separate customer groups, and they may break their advertisement methods down to a few to use it on those customer clusters.

  • A business question that could be answered through the use of clustering would be if a company wanted to look at the specific demographics of their customers, and use special marketing towards groups that are alike. For example, Target may want to market towards stay-at-home mother’s who will be home during the day, and will shop during more low volume times where they would like to see an increase in sales. Clustering could group age, sex, occupation, and likeliness of shopping during the day to find these customers and market to them specifically.

Leave a Reply

Your email address will not be published. Required fields are marked *

Where and when do we meet?
Alter Hall 232
11:00 - 12:20 Tuesdays and Thursdays
Office Hours
David Schuff (instructor):
10:00-11:00, Tuesdays and Thursdays
Speakman Hall 207G and email (see my site)

Nodir Zakhidov (ITA):
Monday: 1:00-2:00
Wednesday: 1:00-2:00
Speakman Hall 207 and email (see his site)