Derek Steer is co-founder and CEO of Mode Analytics, the most collaborative platform for business intelligence and interactive data science.
Twenty years ago, critical questions about a business — such as how to increase the number of sales leads in order to grow — were largely answered anecdotally, based on the beliefs and experiences of people in key leadership positions. Though many companies would’ve preferred to use data when making major business decisions, the process of collecting and analyzing it was too complicated and expensive for many organizations.
In the past two decades, we’ve seen major advancements in data technology, with the introduction of a slew of new tools to help with data logging, ingestion, storage, transformation and consumption. As a result, we can easily and affordably track more information about all aspects of the business — support tickets, orders, web traffic, product usage, market segmentation, customer profiles and more. At this point, most forward-looking companies are already in the habit of looking at data to determine what’s happening with key business functions (e.g., revenue operations, marketing, demand generation, customer success). Organizations that aren’t using data to drive decisions for these kinds of business functions are at a distinct disadvantage.
But data itself, like the raw oil it’s often compared to, isn’t useful on its own. Companies hire analysts to extract information from their data, but a team of analysts isn’t enough. Data analysts are not intended to serve as a help desk that dispenses answers and charts but rather as an internal resource that partners with business stakeholders to ask the right questions. Without support from those partners — without those people asking the right questions of analysts — businesses won’t find the key insights in their data. To best work with analysts, you should start thinking like one.
Channel Your Inner Detective
In most cases, the analytical process begins with the identification of some kind of motivating problem in some area of the business. Data analysts then start to ask a series of questions in order to get to the root of the problem. Most often, they begin with two key questions: 1) What are we really trying to address here? and 2) For what purpose?
Once those initial questions have been answered, they are often followed by a series of additional questions to help the analyst better understand the contextual dynamics of the problem and get an accurate picture of what’s happening.
Let’s say, for example, that Company X wants to know why it doesn’t have more sales leads. This might immediately trigger a sequence of questions such as:
• How long has this been happening?
• Were our leads sufficient before, or have they always been too low?
• Is this something seasonal that happens every year?
• Are leads down across the board, or are they increasing in some market segments and decreasing in others?
• What else might be going on in the market landscape that might be contributing to this?
To answer these questions, the analyst will consult data, looking for anomalies — anything that seems different or surprising. This is analytical reasoning, and in many ways, the process isn’t unlike surveying a crime scene to piece together what happened. Anything that looks out of place is a potential clue. That’s why it’s so critical for business stakeholders to be analysts’ partners — they know the business and can tell when something seems amiss.
Once the analyst has achieved a solid sense of context, they will further investigate in a number of different directions until they can formulate a hypothesis about what happened and why. For example, using the same example above, the next set of questions might include:
• Are our marketing dollars successfully reaching the right market segments?
• Is our product still relevant to the target customers in those market segments?
• Do we understand how those target customers are using our products, and what features are most important to them?
• Is our product being priced appropriately?
• What other products exist in the competitive landscape?
These questions will lead to additional questions or avenues to pursue until the analyst can form a hypothesis about what happened and why. Finally, the analyst needs to come up with ways to test the hypothesis and see if they can find evidence to support or refute it, adjusting as they go.
The key to discovering the truth is digging until you find the answer you need. Here are three tips for approaching business problems like a data analyst:
1. Don’t follow a formula. In order for data analysis to be successful, it’s crucial that it doesn’t become formulaic. The process needs to fit the problem at hand, and one formula won’t fit every problem. Rather than following the same trajectory every time, be observant and curious, and continually ask what’s missing.
2. If you see something, say something. The role of the data analyst is to constantly look for things that seem out of place and push themselves to ask why that might be. If something seems surprising, it needs to be further investigated, not pushed to the side.
3. Iterate as you go. Instead of asking questions in a predictable, linear fashion, data analysis often follows a more meandering, creative process: We ask a question, discover our first effort to answer is incomplete, reshape our understanding by bringing in new data, rework our hypothesis, adjust our questions and so on. This process, referred to as iterative analysis, is central to answering the undefined questions that sit at the heart of any big business decision.
Finally, remain vigilant about staying open to new possibilities as you integrate additional information. Keep going until you reach a reasonable degree of certainty that you have an accurate understanding of the problem and have determined some clear action, or set of actions, to be taken. You’ll know you are finished when you have a more compelling business problem to solve or a new set of questions to answer.