After joining several sharing sessions with UC Davis M.S. Business Analytics(MSBA) alumni who are currently working with Tesla and Facebook, one of the major behavioral questions they’ve encountered was “How do you deal with ambiguity and uncertainties?”. This problem can also be formulated to:
“Can you make the best decision given the limited amount of information and resources?”
Since the topic itself sounds ambiguous and broad, this blog will be centering around the solutions that we can adopt when we run into these types of problems. In the analytical world, ambiguity could come from getting a large amount of data in-hand and couldn’t find the right direction to start with, or it can be that we don’t know how to relate our project to meet our client’s needs. Hence, here are some practical ways for you to consider.
After working with BlueMatrix, our UC Davis MSBA practicum partner, for almost four months, we have gone through the process of giving ambiguous project description, to successfully set up clear instructions and goals to follow along for our team.
“What can we do with the available resources? “ is the question that first came to my mind after getting the project proposal. For research distribution companies, one of their main goals is to manage relationships between buy-side and sell-side firms. Given the complexity between research distribution channels and different needs from the customers, we have landed our first major project, which is called the “Readership Benchmarking.” .We were expected to help Blue Matrix analyze reader’s behaviors, analyst trends, or even market trends to help the business acquire more customers from the globe.
How does the Project Fits into Business Goals
Working on a project that can make an impact on the business is important. Our team started with conducting industry research and company research, to better understand the current trends in the industry. Blue Matrix has more than 38% of the market share within the U.S. and they expanded their business into Asia, Europe, and more. The other 62% of the market share came from the “Bulge Market” including investment companies such as J.P.Morgan and Goldman Sachs. After knowing the market trends and their major accomplishments, we have a better understanding of the direction that we are going which is to help them acquire new customers within the U.S. In addition, given that this project could potentially facilitate their revenue, one other aspect is to increase customer benefits through our analytical dashboards. One example is to provide analyst ranking and article rankings so that analysts can easily find their positions in the market and whether or not to write more on certain topics/industries. Having a deep understanding of how your project could align with business goals is the first crucial step.
Asking the Right Question
With the limited amount of time that we can talk with our industry partner, talking efficiently becomes inevitable during that 1-hour meeting. In order to know when and what questions to ask, we need to do some pre-work. As mentioned in the article written by professor Andrea Follmer Greenhoot from the University of Kansas, she addressed that by analyzing datasets that have very large samples and different measures of constructs, researchers will be able to address additional questions that previously haven’t been noticed. For our project, we did several rounds of data exploration just to make sure that our questions have no redundancy and must be answered before we move on to the next step. In addition, we asked clarifying questions to the technical team to understand the dataset and logins and asked the Sales & Marketing team to understand the critical problems they would like us to solve, which helped us to narrow down the project scope into more practical steps. During this process, we have also identified the pinpoint that would substantially affect our project timeline. The dataset required an extra amount of cleaning to identify the same readers who signed up with multiple email accounts, as well as identifying the industry to which a stock belongs based on the GICS file.
Consulting with Professionals
When there’s a new topic that’s involved in the project or it requires advanced knowledge on a certain tool, it’s always necessary to consult with professionals in that area. Our project requires us to use Machine Learning algorithms such as Natural Language Processing to unify readers, sectors, and stock tickers. We talked with our technical supporter, to gain further insights on calculating reading sessions for each document, based on different sessions recorded in the dataset. We also understand how to integrate NLP into our model, which helped us with the model building from scratch.
Learning how to convert the ambiguity into real business problems can be difficult as each case is unique and requires people to take different actions. However, the essentials are common among all situations and we should be able to narrow down the ambiguities we have into practical steps that we can start to work on.
Reference：
Secondary Data Analysis: An Important Tool for Addressing Developmental Questions. (n.d.). Retrieved January 22, 2021, from https://www.tandfonline.com/doi/full/10.1080/15248372.2012.646613?casa_token=ITNc8v1RlVIAAAAA%3ALeBTskUg3_kTr1KkSkfKCUwnBs_hpwfK2UYPFEf4HZ5NCxsVa2oOj1QjKZZZ5NDGNYd2DifZRg
