Dashboards are integral to any company as it provides data that can be leveraged into business decisions from the insight garnered from it. Part of my role on the DAM team was helping to reconcile the raw data as it came from the application with a set of controlled vocabulary designations for the dashboards. A major challenge in this work is attempting to standardize under the protocol of a singular corporate entity, the varyingly unique structures of its subsidiary brands, while leaving room for each subsidiary brand to retain certain aspects that make each of them unique. It can get even more complex when attempting to do this in the scope of metadata, where subsidiary brand A and subsidiary brand B may have a wholly different way of looking at an asset, or the way either would like to structure their folder taxonomy.
While the challenge is certainly not unique to any corporation, it can be most especially felt in the world of metadata, and dashboard reporting tasks. What helped me overcome this challenge of reconciling multiple, brand-dependent, metadata reports was to first concede that these differences would always exist, second realize the number of possible ways to tackle the issue, and third, understand when to use a particular method for a particular issue.
For example: there were several thousand unmapped assets for a particular brand under one designation, under the asset type category, which was the primary mapping column for the dashboard to map to. The secondary mapping column for these assets were homogeneous as well. The simplest thing to do in this case was to map it to the appropriate controlled vocabulary designation used for the dashboard. In this particular case, this was the best solution because it required the least amount of time to do, while still providing the correct output.
Another example also had several thousand assets with a single designation but differed from the previous case because while the assets were united under the primary designation, they were all different media assets and therefore different under the secondary designation. As a result, the solution was not to singularly map the asset type to a singular corresponding designation from the controlled vocabulary set of the dashboard, but eliminate the primary designation altogether uniting the several thousand assets so the dashboard logic would instead map to the secondary designations, which accurately referred to each asset’s MIME type. Because the dashboard mapping for MIME type was correctly configured from the outset, this was the next best course of action to take. Of course, the caveat to this interim solution was that the elimination of the primary mapping designation was to happen on a recurring basis, given that the raw data being extracted was automatically providing the inaccurate primary designation.
A key to success when being confronted with these automation challenges is understanding the lay of the land. In my case, I was able to identify the strengths of the various teams tasked with the dashboard reporting effort and knew where to go for what. The offshore team’s responsibility was in extracting the reports from the system itself, and so helped me in the producing the metadata reports that would reveal data inconsistencies for the onshore teams. The analytics team’s responsibility was providing the mapping architecture that would transform the raw data into insight for the business teams, and so helped me understand the mapping logic underlying the translations. It was through the combined efforts of our cross functional group that we were able to provide business value for our company.
When done correctly, dashboards for a DAM or any content related management system can provide valuable insight to business teams on the most relevant metadata metrics. Both recognizing the nuanced differences in your metadata mapping process and responding to each different scenario in the most efficient way will go a long way in streamlining your workflow and maximize data output accuracy in your dashboards.