Licking the Data Spoon: Leveraging Insights From Existing Data Before a Digital Transformation

As we approach the holiday season, I recalled one of my favorite activities as a kid: helping bake holiday cookies. The best part was being able to lick the mixing spoons, an early tase of the treats to come. Similarly, digital transformations are in progress or being contemplated by nearly every organization, looking to drive efficiencies and uncover new insights into their business, these businesses are “licking the spoon”. While their promise is strong, there are opportunities that exist today for these organizations to utilize advanced analytics tools to get a taste of post-digitization life before adaption. This sounds great, but there are tools and techniques that work better in this environment than others, plus precautions and prerequisites that must be considered before that first taste. Even more so, these approaches can really be a contributor to EVA (covered more thoroughly in this newsletter) as the analytics can yield operational improvements and profit generation.

A pre-digital transformation organization still collects a significant amount of data that can be used in advanced analysis. Traditionally we have found two hurdles to using this existing data to its fullest potential. First, data controls are not always in place, and second, data exists in various, siloed areas. Poor data quality has always been a limitation of applied advanced analytics, but this is exacerbated in a pre-digital organization for several reasons. The primary factor is a lack of input controls. Open text fields serve as a stop gap to collect data based on someone’s desire at one point of time. Open data fields lead to multiple data issues, including non-standard and inconsistent entries. An example of this would be a customer name that can be entered as both ABC Co. or ABC or ABC company. This issue can be solved through several techniques, like fuzzy matching, using a controlled data field as a proxy (i.e., shipping address instead of customer name) or a mapping of non-standard field data. The key takeaway here is that while data may not be as good as it would be under future circumstances, creative solutions exist to use existing data in advanced analysis. Issues only exist when this is ignored, and erroneous outputs and insights can result.

The second issue, data in various areas, is another reason organizations aren’t taking advantage of advanced analytical capabilities. This issue has been resolved in some respects by the adoption of data warehouses, but some of the most interesting data sets are not generated internally. Historically, tying data sets together has been cumbersome both by a lack of a data key (transaction number, order number, etc.) as well as the tools available to connect the data sets. Excel certainly has its uses but, when it comes to connecting large data sets (the preferred method of non-data scientists), it has its limitations. By utilizing tools like Python and independent databases, it is now possible and efficient to tie together multiple disparate databases.

Recent Experience

SITUATION: An eCommerce based distributor was seeking to reduce their SKU count on their primary site to reduce the cost of maintenance and improve site navigation. The management team had performed an initial analysis but was concerned on the SEO metric impact from cutting SKUs, i.e. Low volume sales items might be driving site traffic.

WHAT WE DID: SLKone, in collaboration with the clients marketing team combined sales and order data from an internal ERP with Google Analytics data on number of sessions, time of session, search results, and click through rates.

THE RESULTS: Combined data set allowed for a unique analysis that accounted for volume, profitability, and site traffic contribution; 60% of the lowest performing products could be rationalized while impacting only 10% of sessions and 2% of users.

A commonly held belief by management teams and organization sponsors is that advanced analytics and the insights they drive are something the organization can only use and achieve once a transformation has been performed. The tools and processes to perform advanced analytics exist and can be applied to the organization right away to identify unique insights. Keep in mind, however, that these analyses are highly customized to the situation and are often difficult to replicate within the organization, so this is not an alternative to transformation. However, the insights gained and obstacles overcome performing this type of analysis can point the way for transformation initiatives. Organizations that don’t wait to utilize their data to the fullest will gain a key advantage in building enterprise value.

Reach out to learn more about SLKone’s application of advanced analytics and the impact they have had for our clients.