In Information Architecture, Modeling

As you may have guessed by now, The Understanding Group loves using visualizations, large and small. They are one of the secret weapons in our never-ending quest to help people make their websites clear and useful.

But visualizations raise all kinds of challenges, not least of which is that they take a lot of time to create. If we are trying to provide maximum value to a client, do we always have to go the whole hog with detailed, time-consuming charts or graphics?

This is a constant source of tension in analysis, and to my knowledge there are few heuristics for knowing how much is “enough.” Generally speaking, it’s based on the pull of the person doing the visualizations and the size of the project.

But it shouldn’t be like that! Using the two simple rules I present below, you can determine how to present information at the appropriate level of detail for any project.

Rule I: Use the “Iron Triangle”

Use the “Iron Triangle” of visualizations to determine which part of the stack you need for your audience. Software development often describes the Iron Triangle of software: in a world where you have software that is made quickly, cheaply, or well…you get to pick two.

There’s a kind of iron triangle of presentations as well: In a world of where you have a presentation that is created cheaply, tackles complexity, or speaks to a broad audience, you get to pick two as well. Whichever two you choose will then guide the kind of visualization you should plan on creating for the project.

So which one do we use? Well, first we should learn a little more about the three basic kinds of presentations that exist. Broadly speaking, data can be presented in three ways:

  • Tables and raw data—also, screen shots from the data analysis tools
  • Narrative text
  • Charts and graphics (traditional data charts, other visualizations)
  • Visualization Stack

    Tables and raw data are the tables that the analyst often is looking at to gain their own understanding. In some cases these can be the actual notes from a meeting or a spreadsheet downloaded from an analytics tool. In some cases, there is a lot of useful information that can be spotted at this level, but the risk is that something important might be missed or that the information will be confusing.

    Narrative text is the analyst’s written description of what is going on, in basic paragraph form or outline summary. Basically this is the analyst telling the audience what they are seeing in the findings and providing some interpretation.

    Charts and graphics are visual representations of the raw data, guided by the analyst. Often they are used to tease out a key point or help explain a complicated concept.

    As you can imagine, effort for presenting each model is relatively low for raw data, somewhat higher for a narrative, and considerably higher for charts and graphics.

    Returning to the Iron Triangle, then, which priorities the project has helps you know what kinds of visualizations you should create. Is the information complex and budget tight? You will most likely be selecting raw tables, but don’t expect them to make sense to anyone outside the immediate team. Do you need to broadcast the findings to a major segment of your client’s organization? Charts and graphics are probably the way to go, but make sure you set aside budget to make them.

    Visualization_triangle

    Rule II: Plan

    Planning your effort and budget based on one of the three basic information models is all well and good, but a project manager will still want to be able to plan effort and cost for visualizations. That turns out to be surprisingly easy. If we look at a typical effort for creating information models, the distribution looks a lot like this:

    • 10% elicitation and understanding
    • 10% data pull
    • 20% data cleanup
    • 20% analysis
    • 10% narrative
    • 30% visualizations

    These efforts can be added/removed from project scope, but only in a specific sequence. Moreover, generally speaking, if you get one of the less expensive items, you get to keep it as you move up the scale. So, a person creating a narrative will also be able to show the raw data and tables; after all, they needed them to write the narrative, and anyone who has charts and graphs should have a narrative as well, because they probably needed to write it in order to determine what they would present in the first place.

    Visualization Effort

    This approach can help understand planning and scope, but the most important question to ask about any presentation of information is: who is going to be using the information? Will they need a more compelling distillation of the information?

    So, as you start thinking about your next presentation, apply the Iron Triangle and estimate distribution to your planning, and see what comes out. It may save you time in some places that you can use to provide deeper analysis and value where it is more urgently needed!

Recommended Posts