Final Blog

After completion of all of the modules for this course, I realized that the modules are designed to support one main concept and that is to harness the proliferation of data in this Big Data era, to generate knowledge and to create value for organizations. Business Intelligence tools can be applied to different industries and domains to provide actionable insights and lead to better decision making. Examples of this can be for the medical field to understand the spread of cancer, forecasting trading opportunities for stocks, or to better enhance content on an e-commerce platform to retain the customer base. The types of information that one can derive from different sources of data are endless and organizations can tailor their data analysis based on questions they would like to find answers to. In order for a business to gain a competitive advantage in the industry, it must leverage Business Intelligence to gather and analyze the data to set up goals that align with the business initiatives. Then, the business should implement and monitor the metrics to ensure that the performance is in line with the target. Lastly, since industry rivals are always playing catch up, the business must continue to innovate to create a new value proposition for its customers in order to remain a differentiator in the market and gain a period of competitive advantage.

Introduction to Big Data

During the first week of class, we were introduced to Big Data and how "Big" represents the high variety, large volume, and high speed of the data that are being ingested through all different types of sources such as IoT devices, social media content, our digital footprints, and many others as we live in a world of datafication. According to IBM, there is also a fourth V to Big Data and that is veracity or the quality and accuracy of the data. Veracity discussed the importance of collecting accurate metrics that really matter and recommended using native APIs when possible. Although veracity is not explicitly discussed in this course, I believe it is always important to assess the quality of data to avoid duplicates or unmeaningful data, which supports the Data Quality Analysis lecture of this course. With Big Data where the data population is N = ALL, it opens the doors of opportunities for different fields to make sense of all the data that's available and to use BI tools to convert data into knowledge and to wisdom, which allows the organization to continue to improve.

Data Warehouse

After the introduction to Big Data, the data warehouse, and its design cycle was discussed to understand the difference between OLTP, Online Transaction Processing, and OLAP or Online Analytical Processing. The difference between the two is that OLTP is usually for databases to perform lots of reads, writes, and updates like a salesperson inputting an order for example. OLAP, on the other hand, contains data across a period of time and is used to support management decision making. Data warehouses are of interest to us since it is a significant tool for business intelligence. An example of OLAP would be an online store trying to understand which product generated the most revenue over the course of 6 months. There are two types of dimensional modeling to design data warehouses, star schema, and snowflake schema. The snowflake schema differs from a star schema where it normalizes the dimension tables. Nevertheless, a business should decide between the two models based on their business needs, nature of the business, cost and time that they are willing to invest in, and also the levels of dependencies between functions.

The first assignment to design a data warehouse was very challenging for me because I was not able to physically interact with the information on an actual database. I also found it hard to come up with categories of questions that can be answered using the data warehouse. Even though I have learned about degenerate dimension, role-playing dimension, and time dimension through the lectures, it was difficult for me to understand how to manipulate the check-in time, check-out time and as well as the procedure times. Aside from that, I followed the four-step process to design the fact and dimension tables by selecting the business process model, declaring the grain, choosing the dimensions and finally choosing the facts. Upon completion of this assignment, I was able to better understand the difference between facts and dimensions and how a time dimension can be shared with multiple role-playing dimensions. I also learned that it is super important to declare the different categories of questions ahead of time, since it will serve as a basis for the design of the data warehouse.

Balanced Scorecard

During the first module of this course, we also discussed Balanced Scorecards and how it is very useful in linking organizational goals to the strategy that's needed to achieve these goals. Balanced Scorecards assist in identifying metrics and KPIs to measure how an organization is performing against its goals. Balanced Scorecard focuses on four different perspectives of the organization including financial, internal business processes, customer, and learning and growth. With a Balanced Scorecard, a business can focus on applying strategies to all four perspectives and set up goals to improve each of these perspectives such as excelling at customer service for the customer perspective. The course also provided an excellent example of how Southwest Airlines successful developed strategies using the Balanced Scorecard and assigned objectives, measures, targets, and initiatives to each of its perspectives. However, it is important to note that in order for a Balanced Scorecard to succeed, it should be kept as a living document and all departments should understand and agree upon the vision from the Balanced Scorecard.

Designing Dashboards and Tableau

I found the module regarding dashboards very interesting as it emphasized the importance of less is more. I understood that just like any other business intelligence tools that we have used previously in this course, a dashboard should be designed to reveal a specific topic of interest for the end-user, at a glance. Dashboards like Tableau is an excellent data visualization tool for businesses to make important decisions without having to combine tons of excel files and waiting for the IT team to generate a report. Dashboards should also not be cluttered but should rather allow instant detection of any abnormal values to allow further drill-downs.

Assignment II. allowed me to again come up with my own questions and suggestions based on the bird strikes data. After my previous experience with Assignment I., I realized that I was able to derive questions regarding this data much easier this time. When interacting with Tableau, which I have never previously encountered, I felt that it was very user-friendly and I was able to directly drag and release the different attributes to create instant visualizations without having to perform complex joins. I feel like it would also be easy for those without a technical background to conduct Tableau dashboards and use it to make key business decisions. I also enjoyed identifying the answers to my questions such as when and where are bird strikes typically happening. Previously, I did not know that there were so many bird strikes occurring in the U.S. that caused a significant cost to repair the airplanes.

Web Metrics and Google Analytics

Week 5's material was my favorite module as we explored Web Analytics and Google Analytics. Web Analytics allows businesses to better understand the visitors to their websites by asking the questions of the Five W's. Other metrics for Web Analytics includes the number of visitors, exit rate, bounce rate, conversion rate, and task completion rate. There are also different types of web traffics such as direct traffic, organic traffic, referral traffic, and campaign traffic. A business can look into the different types of web traffic and their corresponding web metrics such as conversion rate and bounce rate, to determine if they would need to fix the landing page for a type of traffic or invest in more money in traffics with a high conversion rate and low bounce rate.

While I was experimenting with Google Analytics using the Google Merchandise Store for Assignment III., I was so amazed that one could generate so many types of reports from GA such as the audience report, behavior report, conversion report, and also custom reports. It never occurred to me before this class that when I was exiting a site, it would be recorded as a bounce rate on their platform. By using Google Analytics, a company is now able to better identify their target users, what types of content they are interested in browsing, and when are these users most active. An organization can set up its goals in Google Analytics and measure their conversion rates and how they can continue to improve their sites as they began to tailor the site to fit the users' needs. An excellent example of an organization using GA is an e-commerce website that is looking to identify the demographics of their visitors, the merchandise that generates the highest revenue, and as well as which page has the highest exit rate. If the page with the highest exit rate is the checkout page, the business should look into ways to implement a more convenient and appealing checkout page.

Network Analysis

For the last module of this course, we discussed the network and how we can use network analysis and network visualization to understand very complex systems and their behaviors. With network visualization, it allows fast detection of communities and how the members are interacting with each other in the network. It also assists in differentiating between two networks that otherwise have very similar network statistics such as in the case with Google vs Apple Patent Analysis. Network analysis and visualization can be used in almost any scenario that requires connections and interactions. Some topics to conduct a network analysis could be on Linkedin members and their relationships with one another, a disease that's highly contagious and spreading across communities, or it could be the crime in a neighborhood with critical members that are facilitating the crime. By using network visualization in combination with the different network properties such as betweenness centrality, density, and Eigenvector centrality, it allows the analyst to see the full picture and get a better understanding of the complex relationships between nodes and to act and plan accordingly to achieve the organization's goal. 

The international trade data of 2017 represented an excellent source of data for network visualization and analysis using Gephi as international trade has always been an extremely complex phenomenon. Since I did not have any previous knowledge about international trade, I feel that this assignment really broadened my horizon and killed two birds with one stone. By visualizing the network of trade between countries, I identified different communities on Gephi and ranked the size of the nodes based on their betweenness centrality. The observations that I have made from Gephi aligned with my readings about international trade where countries in the same communities have a better trade relationship than those outside of their community. The countries with the highest centrality values demonstrated the greatest influence in international trade and had higher values of exports and imports compared to other countries. Since I have previously heard about Brexit, this assignment also provided a great opportunity for me to identify growing trade partners for the United Kingdom as they will no longer receive free trade with members of the EU after the divorce. I'm excited to see what's going to happen to the United Kingdom after Brexit and if my predictions are correct!

Final Thoughts

Overall, I really enjoyed this course! Although there were challenging assignments that required me to work with completely new BI tools for the first time, it also pushed me to pick up these skills and feel more confident in designing solutions for businesses. I also felt that I was able to think more critically by constantly brainstorming questions that businesses would like to seek answers to, and use Business Intelligence tools to answer these questions and make critical decisions. After finishing all the modules of this course, I realized that the course has taken me through many phases of the BI life cycle, including Visualization and Descriptive Analysis, Reporting and Dashboards, and Determining Questions. My final recommendation for organizations out there is that, why not take advantage of this massive amount of available data and apply Business Intelligence to innovate, improve performance, and become a differentiator in the market?

References

Foster, Patrick. What Are the 4 Vs of Big Data & How to Apply Them?. Retrieved from https://forwardleading.co.uk/blog/What-Are-the-4-Vs-of-Big-Data-How-to-Apply-Them.

Jackson, Ted. How To Use The Balanced Scorecard (& Know If Yours Is Failing). Retrieved from https://www.clearpointstrategy.com/how-to-use-balanced-scorecard/?fbclid=IwAR1kYhq2LS0sewTN-d8U0TsOVnnCRk7j_1f-GQK6qxTMFsbUeDvoQ6UdkX0.

McAfee, A., & Brynjolfsson, E. (2008).  Investing in IT that makes a competitive difference.  Harvard Business Review, 87(7/8), 98.

Sansu, George. Inmon or Kimball: Which approach is suitable for your data warehouse?. Retrieved from https://www.computerweekly.com/tip/Inmon-or-Kimball-Which-approach-is-suitable-for-your-data-warehouse?fbclid=IwAR3Nq8u4YeZE1oDRrUMbm6tL7d7CNMLx_A8LlaxlV8N3u4ggtnanFcUyB3Q.