Big data is no longer a single-industry term. In fact, if you work in the digital arena in any capacity, it’s likely you’ve not only heard of big data, but that you actually use it on a regular basis.
Now, the use of big data requires the successful completion of at least one major hurdle: actual collection. In general, our collection of data has become so sophisticated and ubiquitous that a quarter of data scientists say Hadoop, the big data industry’s big data cruncher, is too limited and doesn’t allow for a sufficient amount of data sources. What’s more is that 71% of data scientists say that the amount of data types, sources and varieties is making their jobs much, much more difficult (Hadoop or not).
In other words, it’s possible that big data is getting a bit too big. For companies and executives looking to spend on a big data solution, this over exaggeration of data can wear out a budget, eliminate any hopes for ROI, or overstimulate a team while producing little to no actionable insights from the information (again, eliminating any hopes for ROI). The solution here isn’t to hire the remaining 29% of data scientists undeterred from the growing size of big data nor is it to give up on big data entirely. Instead, the key is in differentiating a data-driven strategy from data-as-a-strategy, thus determining volume based on need, rather than possibility.
Data-Driven Strategy vs. Data-as-a-Strategy
A 2011 McKinsey study reports the potential value-add of a data-driven strategy as follows:
- 60% increase in operating margins for U.S. retailers
- $300B annual savings for the U.S. healthcare system
- $600B annual consumer surplus that we, the people, can reclaim from successful implementation of personal location services
“There’s really no benefit [of] big data to a business,” says Nick Goggans, Chief Strategy Officer at Umbel, “unless you can query it and get an answer.”
In other words, data just is. If any of us want to gain real business benefit from the data collected, we must actively extract that value by making our data actionable. Only a data-driven strategy will accomplish this. Data alone is not a strategy, nor a solution, nor a saving grace.
A data-driven strategy implies that leadership can intelligently use their collected data points as a multiplier by determining which lever to pull. Teams with a clear understanding of their current environment and an active STP (Segment, Target, Position) strategy can increase that multiple by strategically using data to extract a higher percentage of the potential value-add.
Implementing a Data-Driven Strategy
First, business leaders must decide what real business benefits they want from the data before determining which questions to ask of it. Keep in mind that more specific questions lead to more actionable answers. For example, knowing you want a high ROI is not specific enough as the data is too big to address ambiguous goals. Aspiring analytics leaders need a deeper level of specificity and originality than “increase customer spend,” or similar ambiguous goals may be.
Second, the multiplier effect comes by way of clear decision nodes and modeled outcomes of each decision. In addition to sales force allocation, parsed and properly queried data can reveal:
- Which SKUs to stock
- When to run a campaign or promotion
- How to allocate MSA (Marketing, Sales and Advertising) spend across customers, channels and products
- How to create a more seamless customer experience through improved after-sales service
- How best to contact each customer segment
Thoughtfully parsed data allows companies to attack one piece of information or a single goal from several different angles, increasing the success rate.
For example, in order to understand what drives online reviews of a specific product, we look at which customers are the most active reviewers, which products are the most reviewed, and what the correlation is among order frequency, product price and number of reviews as well as the rating of the review. Looking at these layers can offer insights and connections otherwise missed by a shallow analysis. By comparing the end values along each branch of the decision tree and looking at cross-sectionals of where parsed data intercept, we can effectively code the qualitative aspects we want to capture. These predicted outcomes combined with outside market environments give us a robust model to work with – and a deeper understanding of our customers.
“There’s really no benefit [of] big data to a business,” says Nick Goggans, “unless you can query it and get an answer.”
In the end, using data as a multiplier, or utilizing a data-driven strategy, is all about figuring out customer lifetime value and strategically increasing that number across customers. No longer do pageviews determine the success of a digital company. Instead, user engagement and lifetime value reign supreme. These metrics pull in sponsors, create niche, loyal communities and allow companies to thrive based off of a customer service that caters to a customer’s ultimate desire: convenience and discovery.
In marketing, they say that to delight and surprise is how one keeps its best customers. In the ever-growing Internet of Things, it will be data-driven strategies that determine where companies spend, all while decreasing risk.
Don’t fall into the trap of data-as-a-strategy. After all, the data merely reflects individual points of interest. People are behind those 1s and 0s. People need to be the ones querying it and intelligently (albeit ethically) using it to boost their business as well.