Did you find a solution for this? There is a level of data manipulation that needs to be done first and a few other questions that I would have.
For instance, most of the accounts reference a single year. Are these excluded from this analysis?
What about Product types? most accounts with multiple references across years, contain data for a single product type. So oen might be X value with product type 2 in 2011 and Y value with product 7 in 2012. Most accounts with values for different years also have different products.
Do accounts with no values for other years become equal to zero? for instance, should we treat 2013 null values as zero and include them in the calculation of the 10% or entirely exclude them?
A bit of understanding how these product types and year differences play into your situation is helpful to determine the top 10% and what to exclude is needed.
Also, do you have a Tableau workbook showing what you have tried so far? This helps the community to better understand where you are at with the issue.