How a person receives a product is as important as the product itself, with truck drivers being a great way in understanding this point.
In the US, up to 72% of freight by value is delivered via truck. These products come through a combination of boats, planes, and or trains, but at some point a majority of them are driven to their final destination. Let’s say trucker drivers were uncharacteristically inconsistent and you were a small business owner. Could your business grow in this made-up system? Maybe. Growth would be difficult because no matter how good your product was it might not get to the right person in time, souring customers. The same is true for our designs. It’s not enough to have excellent data visualizations—you need to eliminate barriers to access by understanding the user’s journey.
In some cases a user’s journey is a simple search for information and there is ample research on how people search for information. The interest in studying information searching makes sense given the amount of choice in our world. That is, there are so many options or details for literally any product that users have to make more decisions overall in their journey. In fact, there is a whole field of study on the subject of a user’s journey—user experience (UX) and user interface (UI). Designing a top tier journey for your users, and not just their interaction, requires some level of expertise in either UX or UI. That said, I believe as designers we can still make better choices while we develop that understanding or connect with those experts.
Our contextual discussion so far has considered truck drivers, toothpaste brands, and computer science research. By any measure that is a lot of disparate topics, even for one of my articles. This is why I think it helps to review a more realistic example; an example that is based on my personal experience. The phone rings and it’s a prospect getting back to a decision maker ahead of schedule. The decision maker remembers this amazing dashboard you shared in a recent meeting and wants to share some insight from it. Unless they have a printed copy on their desk, the decision maker needs to locate your data visualization on their computer before they can study or share it. So, how does the decision maker find your creation? What steps do they take? What is our user’s journey?
In my experience at for-profit, not-for-profit, government, and academia, I have seen four major pathways to how people typically access data visualizations. Understanding these simple pathways can help you ease the user journey for your audience. If for no other reason than an awareness can allow you to become more efficient at these pathways. The major pathways that I have seen in data visualizations are: a gatekeeper, a file structure, a saved list of links, and a keyword search. Most organizations house more than one pathway and often their development is more organic need-based versus mindful systems-based.
A gatekeeper is a person or AI with which the users interact to get the content. This is the oldest pathway I know of and the only one that exists in every job I have ever had, to varying degrees. In our above example, the decision maker would not attempt searching for the content. They would either ask their aid to reach out to you or they would reach out to you themselves. This pathway represents a double edged sword. If the designer is receiving requests for their content, then it is clear the team desires to use that content. It is also clear that internal presentations in meetings are displaying the value of this content effectively. Otherwise they wouldn’t care to call. This is wonderful! The opposite side being that the user may view the designer as transactional versus consultative, calling only the moment they see a pain point. This implies their data visualization literacy is limited to knowing when the content is valid versus knowing the content. Over reliance on this model can create unrealistic time demands on your schedule. This is why most places also have some sort of self-service clearing house of content.
A file structure is a series of folders, real or digital, that create a storage system the users can learn to locate the right content. Abstract or real versions of file structures are in lots of places from the library, to grocery store aisles, to your email inbox, and the local newspaper. Setting these up is harder than it looks. It in fact leads to the question, “how do you group things for users?” There is alphabetical sorting, which is very common due to the shared knowledge of the alphabet, but challenging given it requires the person to recall the name of the item. The decision maker who just got the call, may not exactly remember your title. This is why I tend to group my visualizations, when using this pathway, into non-overlapping categories each with its own definition. That way the decision maker need only recall that it was about “board meetings” and they can flip through the content in the folder. Definitions are key for the teammates understanding as they become references, but also for your own classification purposes.
A list of saved links works like a favorites list from your web browser or a playlist from Spotify. In fact, here at The DeBruce Foundation, I refer to Tableau’s Collections as Playlists, which have been very helpful in cutting down the amount of content the user needs to process in their journey. There are also plenty of users who keep a list of links on their web browser. It’s hard to consider an analog example of saved links, perhaps the drawer or physical location? Regardless, the challenge of this pathway is also its strength. What the user needs is one click away. The focus allows for a great deal of speed and comfort; however, saving a link is something a user will only apply to known content of importance. New reports will not automatically populate a user’s Google Chrome favorite list and designers may struggle to introduce new content to their users. With a file structure, for example, when the user opens up the “board meetings” folder they can see any new additions.
This is where Tableau’s Collections really stand out. As an administrator to Tableau Cloud, I can add new reports to the user’s playlist on their behalf. It is important to gain approval first, but providing “white glove” experience in list management is something that shouldn’t be understated. If your clearinghouse for reporting has the ability to create user specific lists, then I would suggest it is one of the best pathways here.
A keyword search involves users searching a website, file structure, or computer by inputting words or phrases associated with the content to locate that content. In a way this pathway is a more abstract version of the file structure with the keywords acting as folders grouping different items. Interestingly this function exists outside our design programs! Right click on any Microsoft Office file or JPEG and select the “Details” tab. From there should be a “Description” section. I believe at least the “Tags” item is searchable by file exploring software. This means you can apply this to your own internal documentation, not just in design interfaces like Tableau Cloud. The challenge here lies in which words become keywords. Select the correct words and the search process feels intuitive to the user. Select the wrong words and the staff now has to memorize a series of random prompts for them.
This pathway is by far the most powerful and the most complicated. Thanks to search engines’ popularity almost all of us know how to search via a search bar. This means teaching your teammates how to find designs is much more about pointing them in the right direction than introducing technology. This is part of the answer, but it does not yet address the question of which tag or keywords to use. Again, I am sure there is deep analysis on this subject, but for me I break this into two steps.
Step one, decide on a table of contents index of keywords or more free form. Table of contents works like grocery store signs—they are more or less the same in each store and folks have to memorize, to some degree, whether items fit in a category. This makes perfect sense with limited subject matter. That is classification is easier when there are less unique items to classify. If you have a more diverse collection of subject matter, then creating a universal classification system may be very hard to accomplish. This is why I tend to go the route of numerous keywords tying those words and their variants to different grouping topics. The upside is that there is little memorization required by the user, the downside is that the keywords cannot be used easily as an index.
Step two, decide on the keywords. In either case, index or free form, I think the best strategy is to build keywords around how people search for data visualizations. Below are the search concepts I have used before, these represent the meta categories that associate with a person’s interests. From these several keywords would need to be generated, example keywords are in the parentheses. Again, I must stress non-overlapping keywords and having some definition, even if loose, needs to be in your documentation.
- Type of audience (Partner, public)
- Type of associated department (Sales, partnerships)
- Type of location (US, international)
- Type of goal (Goal 1, goal 2c)
- Type of topic (Board meeting, training)
- Type of data (Survey, web traffic)
- Type of data activity (Comparison, Composition)
- Type of visualization format (Report, slide)
- Type of chart (Area, Bar)
There is no silver bullet for helping teammates find your work. It takes time to understand the pathways already established in your environment and more time past that to master them. The key in any of these situations is documentation and consistency of action. Even in the gatekeeper pathway you can still set up office hours for when people can call you. That said as designers we are used to iteration in design and how teammates access your work is no different.
Cover image is AI-generated.
Christopher Laubenthal focuses on better data use with visualizations in an organizational setting. He has experience in both for-profit and not-for-profit sectors where he increases literacy, grows culture, and builds data visualizations. Christopher is the Data Design Manager at The DeBruce Foundation, a national foundation whose mission is to expand pathways to economic growth and opportunity. Current projects include his public viz and The DeBruce Foundation’s Career Explorer Tools.
