In 2012, Harvard Business Review’s sexiest job of the 21st century was Data Scientist.
It’s no wonder, given today’s never-ending appetite for data. The prevalence and importance of analytics can be seen across all realms of business, and top universities have begun teaching data science as part of their curricula. Even top MBA programs are teaching tools and programming languages to prepare their students. Data and analytics have moved from almost exclusively IT and Finance into nearly every aspect of modern business. But, there is a deeper truth that few people talk about openly: data alone may do little to help your business, and, when handled incorrectly, can be painfully destructive.
The “data-driven decision-making” trend prompted leaders to develop a nasty habit of looking for more data to prove any and all ideas. It became the be-all and end-all of everything. Despite this, the success of “data-driven decision-making” depends less on how much data is available and more on what you can do with it. For this model to be successful, we need to address the following challenges*:
1. Ask the Right Questions
It’s not solely about relying on data. It’s about asking the right questions and making sure you incorporate them at the right time. If necessary, then use the data available to validate or invalidate your hypothesis.
Example: If your conversion rates are going down and average order value is going up, and you want to understand why this is happening, it is essential to map out all the possible scenarios before you look at the data itself. It is possible you are seeing an increase in high-value customers through recently activated channels, or a direct result of a pricing change, or this could be an industry trend or lack of competitive spending. If you do not list all questions before diving into the analysis, you will waste a lot of time trying to identify the issue and miss out on the most important opportunities.
2. Data is Not Insight
There is often a missing link between data and business value. That missing link is generating actionable insights. This occurs at the intersection of business understanding, analytical thinking, and the necessary data to prove/disprove the hypothesis.
Example: You have two different CPAs for two different channels. One is for branded search at $20, and the other is for TV at $100. If you merely look at these numbers, TV at first glance is not very efficient. BUT, if you understand the interaction TV has on branded search’s performance, you will realize that the $100 CPA for TV could hypothetically come out to $60. This context is essential to deriving insights from data.
3. Garbage in, Garbage Out
By now most of us have accepted that no matter what we do, we are in the business of generating data. The reality, though, is most companies do not begin with a clear data strategy in mind, whatever their intentions may be. They usually start with a big idea, rush to market, and worry about cleaning up the data and building a strong foundation after the fact. If you don’t start with the data strategy in mind, you will continuously be playing catch-up, most likely generating what is effectively garbage, and digging the hole deeper as you go along.
Example #1 (the bad): You’re a marketing manager, and you’re using 15 different tools to manage a marketing campaign, which means there are at least 15 different sources of data. Each tool tells a different story, so what ends up happening is you spend three months trying to stitch 15 different outputs together. You lose time and money, and your marketing goals become less and less achievable. Your marketing strategy now becomes one of constant fire drills and catching up.
Example #2 (the good): You’re a marketing manager, and you understand that data comes first. Because of this, you begin by laying out your data strategy in a way that can stitch vendors together holistically. So, before you go ahead and pick the “cool” new vendor, you check if it will fit into your strategy. If yes, you’re good to go. If not, you need to keep looking elsewhere. Once you activate this vendor, you start to see how your whole story moves together without having to create an entirely new project to understand the impact of this new initiative.
In short, data alone will not save you. It can be your most powerful ally, but if not managed properly, your greatest enemy. Without the above three elements, no data-driven organization can produce value. We are in the advertising business, so we have a unique perspective toward companies who have done this well. But from time to time, we also see the results of a poor data strategy and the destructive consequences that undermine important marketing efforts. If you ever need some truthful candor on how you stack up to others in your field, we would be glad to take a look. It can do wonders for your advertising programs.
*This is not the exhaustive list, but the most common.