December 13, 2018
-John Hope - Practice Leader, VP of Analytics
The Early Days of Data Science
In the year of 1996, I was catapulted into my first job as a quantitative analyst for a publisher who was trying to maximize revenues amongst various product lines. We needed more efficient ways to understand customer profiles for our biggest product lines so that we could effectively target them for future campaigns. Direct mail was booming in that pre-Amazon era, however, customer acquisition rates were increasing and as a company, we felt inefficient when it came to reaching potential clients. As a business, we sought more efficient strategies to acquire customers and upsell current customers.
Introducing Quantitative / Data Science
Our team was tasked with taking mounds of data from our mainframe and mining that data to find patterns and profiles of our customers. Frequent questions we had to ponder were: Who are they? What do they buy? and When, and What should they buy next? As business-minded thinkers, we often found ourselves considering the most profitable segments to market. With a trusty Pentium 386 and a SAS licence in our arsenal, we were able to provide insight and visibility into this mound of data and identify segments and targets for our marketing teams, as well as build hefty profiles of our customer base in order to recommend products that clients were more likely to react to positively.
There isn't much difference between the analytics and data wrangling we were doing 20 years ago to now. But the major difference is the ‘time to action the data’ is dramatically reduced now, especially with enterprise platforms like Salesforce. Giving your key stakeholders the ability to consume insight - whether it be from a large data set, or the result of a predictive model, means that your teams can effectively use the data to be more relevant, a task that was almost impossible 20 years ago.
At Bluewolf, an IBM Company, our mission to help our customers become more data-driven in their day-to-day approach of tackling objectives is critical and the process that we follow is listed below.
Gathering data and requirements are necessary, but often one of the most overlooked steps is the accessibility of that information. A great predictive model is nothing more than a solid model unless one can access it, or easily translate where sellers need to focus.
To use Salesforce to your company's competitive advantage it needs to be useful and relevant to your employees. As we found in our own global research report, The State of Salesforce, when it’s easy to use and share data produced by other departments, employees perform better.
Are you ready to tackle objectives with a more data-driven approach? Connect with us.