At Baufest, we are results oriented. We treat our customer’s data in the same way, not to simply build reports or visualizations but an invaluable resource to impact the bottom and top line.
Yes, in larger companies it is very hard for one single individual to make a difference on the overall top or bottom-line, but every piece of data should tie to some positive contribution and we all have our part to play. Therefore, in this time-strapped world, with finite resources, every report, metric, or visualization must speak for itself and say: “this piece of data drove this quantifiable business impact”.
STEP 1. One data-point is interesting, two or more are useful; patterns and relationships are impactful
The first step to introduce is: How to identify potential impact. Specifically, where do you start analysis and not waste time on likely unfruitful analysis. A bonus here is verification that you know your data is accurate.
Just because you have data, does not mean it is all useful or important. There is a lot of information that is “interesting but useless”. For example: log files from a server explains usage of your e-commerce system by the second (or millisecond) – this data point is not that useful unless it is tied to something else. For example, by linking usage in some meaningful way like to server cost optimization, shopping cart value and quantity of items. Once linked, you can demonstrate relationships and effect decisions around costs (server provisioning) and revenue (peak hours of the day when online purchase occur).
Here is a Pro-tip for putting this into action: Start first by thinking about your company’s objectives, goals (measured and time-bound) or key results. Then align the data to these by building some baseline metrics, and measure against numbers you know are correct before going into any complex modeling exercise. Once you put a chart in a dashboard and share it, it is out there. So, make sure you are starting from the right data set to begin with.
Just as important though, you must know the data is reliable and understand how to model it in a way that it produces accurate conclusions. Here, you can test assumptions – if your peaks and valleys do not align with your assumptions, test further. Look for patterns that align to assumptions and anecdotes. Also, speak with someone – take your data to a lead who can verify or disprove your information. I guarantee you will learn something new. I was coached early in my career: when you do not personally possess authority granted by tile or experience, arriving armed with great data leveled the playing field.
STEP 2. Develop a Series of Hypothesis about Your Area of the Business
Once you know the data you are working with and have a good sense of how reliable it is, think about the business as it relates to parts you can control / influence and that aligns to an objective, goal, KPI. If you are in advertising, focus on advertising. If you are in product development, find data sources that tell you something about the customer’s needs. At Baufest, we have seen rogue projects that fail because they do not latter up to a problem that needs to be solved or a goal. Therefore, consider a top-town and bottom-up approach simultaneously. For example: a specific goal to increasing service margin (top-down). Now, what are all the inputs to cost, offering price, fixed and variable elements that can and cannot be predicted. Do you have complete information about both the data AND the processes that lead to an outcome?
Here is a Pro-tip for putting this into action: Start with conversations. Speak with peers, use customer feedback, and gather 3 to 5 anecdotes (or more) that about how the business does operate or could operate differently, then use the data to test out those theories. Anecdotes are almost always right and finding the data to support the anecdote and affect real change is, well, a game-changer.
STEP 3. Focus on Low Hanging KPIs, then Expand from There
It is easy today, with technology making more data accessible with less effort, to try and seek groundbreaking insights. A proven technique to determine where an impact will be made is to start with a sizing exercise. Take a specific goal you are trying to impact, for example when trying to increase service margin, determine the dimensions of that part of the business like revenue, hours, or number of clients for all services in your company and then quantify and organize those dimensions in a way that highlights those differences.
Here is a Pro-tip for putting this into action: Start by creating a Pareto chart of the topic at hand. For example, customer service complaints by volume, revenue by product sold, or transactions per customer. Then, focus on the top few areas that can have the greatest impact – otherwise, you risk working on areas of the business that just are not relevant to making an impact.
You will quickly see that some topics impact the business much more than others. Too often, at Baufest, we see a client work on an area of the business before conducting a sizing exercise. Data can be powerful, but it can also lead people astray or get people focused on the wrong things. By combining all the elements in this simple 3 step process of using validated data tied to a hypothesis and measured in a way that drives a meaningful revenue outcome, it can make a big difference.
Interested in how your data visualization can get fine-tuned? Contact Baufest today at us@bfdevweb.baufest.com to see how we can help super charge your business intelligence development. We have hundreds of years of combined experience building data visualization that drives business change and can help you craft a more meaningful data story.