Throughout high school and college, the above formula for how to write a good essay followed me everywhere I went. The thesis statement defines the argument to be made, the topic sentences break up the argument into smaller points, and the supporting evidence drives it all home. Though my days of writing essays for school have long come to an end, I’ve found that this framework for crafting a strong argument persists beyond the classroom and into the business world. In this post I will lay out the importance of leveraging data in presenting sound arguments, as well as what companies can start doing to get the most from their data.
Making an important business decision intuitively follows the same “thesis > topic sentence > supporting evidence” framework. Consider this example – a company launches a brand new product line, and three months after launch, the VP of Product needs to decide whether to dedicate more resources to the product team. Constructing the argument to expand the product team relies on the thesis that the product has been a financial success and continues to drive value and growth for the organization, while keeping costs as minimal as possible. To support that thesis, the VP of Product might structure their argument around these components (or “topic sentences”):
The product has experienced financial success over the past three months
The target market provides opportunity for sustained growth into the future
The team has demonstrated efficient use of resources
With the argument broken out into its component pieces, the final action of making a successful argument is to back up the “topic sentences” with supporting evidence. In many cases, the best supporting evidence is data. Data allows decision makers to see beyond the scope of their own personal experiences. These experiences are valuable in guiding the argument and ensuring that the right thesis statements are proposed, but data plays a critical role in executing an effective argument.
Unfortunately, there are many obstacles that prohibit organizations from getting access to the data necessary to make compelling, meaningful arguments. A few notable challenges (among many) are:
Data is spread across several systems
Raw data is not always trustworthy
Data does not answer the right questions
Many times, the data required to prove a thesis may be spread out across many different systems. In our product line example, the supporting evidence needed to make a convincing argument would likely pull product financials from an ERP system, customer data from a CRM system, general economic trend data from government databases, and potentially many more sources. For a one-time analysis, it is no problem to pull all the relevant data from the respective systems, but to perform ongoing analysis and monitoring of the supporting evidence that drives the argument, an automated approach to data integration is a must. Decision makers should think about where the supporting evidence required to prove their theses is located and move towards making the data integrated and centrally accessible.
Raw data can often be messy: blank values, outliers, data collection errors, anomalies, and the list goes on. In its raw form, data cannot always be trusted. The strength of the argument depends upon the supporting evidence’s ability to prove the thesis, but if the evidence cannot be trusted as a reliable source of truth, the rest of the argument crumbles. Back to our example, let’s say the VP of Product is using product financials from an operational system to prove the thesis. However, the revenue figures from the operational system do not match that of the accounting system. Before supporting evidence may be used to prove a thesis, it is necessary for the data to undergo a thorough validation process to ensure trust in the data. Decision makers should pay close attention to the quality of the data that underpins their arguments and ensure that data can be a trusted source of supporting evidence.
Though there may be mountains of data in an organization, the true value of that data rests in its ability to answer the right questions. Using the new product line example, if the VP of Product is trying to prove the thesis but does not have data on the weekly and monthly trends of the product line’s revenue and expenses, the argument becomes very difficult to prove. The supporting evidence to prove a thesis needs to be represented at the appropriate level of detail for it to be an effective argumentation tool. Decision makers should start with the thesis they want to prove, then look to the data to identify what may support the thesis. If the data cannot support the thesis, either because it is too detailed, too summarized, or simply not there, decision makers should think about how to procure the data necessary to support their theses.
Thinking about decision making as the practice of crafting an argument to better inform decisions is the first step in realizing the true value of data. Every business fundamentally operates on a set of theses to be proven out, and by integrating data between systems, cleaning and validating untrustworthy data, and finding data that answers the right questions, the organization equips itself with the tools to create compelling arguments to make better decisions and drive real value.