In these lessons, we'll explore the steps an organization can take to efficiently collect data for their projects and objectives. Data collection can be an expensive and time-consuming activity for any organization, so it's important to be aware of the context and purpose for collection and the resources of the organization. When setting out to collect data, it's important to keep the desired outcome in mind and to be as precise as possible when identifying what and how much data is needed. It's unnecessary to collect all the data in the world about a particular subject to achieve a goal, whatever it may be. An organization only needs enough data to reach its desired outcome. As collecting data becomes easier with new technology, organizations may sometimes collect data they think they might need without knowing why they need it. Let's take, for example, a company with multiple store locations looking to collect data to figure out what they should do to increase profits next year. Should they collect data around importing new popular products the company can sell? Collect customer data from each store to figure out which customer demographic is bringing in the most revenue. To figure out whether they would pay more for certain products or look at customer satisfaction data, any of these collection tasks may provide data that can guide how the company could increase its profits. It would be great to gather every data point we can to answer any hypothesis we come up with, but collecting data just in case it could be useful isn't practical. There simply isn't enough company resources and time to indiscriminately collect data for every idea that may potentially increase profits. Even if we start our data collection process by only collecting data easily accessible by the company, we may still end up collecting data that may ultimately have no or very little effect on profits. With that in mind, how can we figure out how a company can increase profits? The first step is to break down what profit means to a company. Profits is a numeric value calculated from the company's revenue minus the company's costs. The formula looks like this, profits equals revenue minus costs. To increase profits, a company can figure out how to either increase revenue more than it increases costs or decrease costs more than it decreases revenue. We can start by collecting the qualitative and quantitative data around revenue and costs for the company. From collecting the initial data around revenue and costs, the organization can then use that data to support and guide the next series of data they need to collect. During the data collection process, the data collector may be influenced by their own personal biases or intellectual curiosities in a way that distracts them from the main objective. For example, a national retailer may be looking to collect data for the objective of increasing customer purchases. First, they probably need to understand who their customers are. A data collector with years of experience in social media may want to collect website visitor and social media data to build out customer online profiles. On the other hand, a data collector with years of experience in customer service might want to collect data from current customers in the store's system and build profiles based on previous purchases. While both approaches may describe the retailer's customers, each data collector has a different preference for the type of data they want. To determine what data is needed, the data collector can go back to the original business objective of how to increase customer purchases for the company. Then, determine what data can solve this specific problem without invoking past experiences. Perhaps a more objective approach is to first understand what the company defines as a customer and what customers are purchasing before diving into data collection. Another potential distraction in data collection is intellectual curiosity. When the data collector drifts from the original purpose of the project, data collection for a purpose can become a pure data exploration exercise. Collecting one data set may influence the desire to look for another related data set. After a while, the data being collected could become unrelated to the original objective. For example, to understand and profile their customers, an e-commerce company tasks an analyst to collect data around distinct visitors to the company's website. While doing so, the analyst discovers that an unnaturally high number of visitors are coming from Spain. With this, he then dives deeper to try to collect data around which regions in Spain the visitors originated from. He then tries to collect data around typical Spanish hobbies from each region to maybe find context on why they could be interested in the company's products. While he continues to collect data that's loosely related to one another, the analyst may end up spending a lot of time collecting data that has little impact on the original objective of providing a profile of the company's global customers.