The three primary functions of tag clouds include allowing the user to easily search subjects alphabetically, support customized prioritization of large amounts of data, and convey the relative importance of various topics on a site. Most tag cloud designs handle data access, business logic, layout and functionality with a single piece of code. However, treating each of these functions individually would allow for greater ease of use of alternatives within the architecture. When creating the source data for a tag cloud, there are three primary restrictions: restrictions on the density of the data, relevancy of the data and timeliness of the data. Separating the source data from the functional layers of the tag cloud is a key improvement. To achieve an even distribution of font sizes, it is necessary to 'linearize' these values enhancing usability over strict accuracy. The Pareto distribution, or '80-20 rule' is one common form of data distribution and a good example of why it is so critical to select the proper algorithm to effect the linearization of the dataset. Regardless of how the UI controls are implemented or the HTML tags are bound to the data, it is worth making the effort to facilitate front end development by using an HTML generator compatible with the most common methods for setting font size, color and controlling hyperlinks. It is possible to create generic solutions with wide applicability through the judicious use of data sources and applied algorithms to determine and distribute font sizes evenly across the tag cloud.