Better Interfaces: Interview, Treemap, Tornado Chart

Innovative user interfaces have helped a number of products get adopted. These are three examples–interview, treemap, and tornado chart–may help to stimulate your thinking.

Better Interfaces: Interview, Treemap, Tornado Chart

1. Interview Metaphor: early tax preparation software automated a literal view of the paper forms. Intuit, I believe, was first to switch to an interview format that captured the information in a way that was easier to understand and manage/update. This was a clever reconceptualization of the problem that has been widely adopted. It was “better” in the sense that it was more easily understood by a consumer who only filled out one tax return a year.

2. Treemaps: developed by Ben Shneiderman compress a lot of information on quantifies that differ by orders of magnitude and are more useful at presenting it than bar charts or pie charts. One of the more successful examples is Smart Money Magazine’s “Map of the Market” which you now have to subscribe to see. Some references

3. Tornado charts: also called Tornado Diagrams, these  are a compact visualization of the impact of different variables on the outcome of a sequence of decisions (or a decision tree).  In decision analysis they obsoleted tables of numbers and probability distribution plots, offering a succinct encapsulation of the relative potency of a number of control or input variables. They have migrated over time from an esoteric technique in specialized software tools to one easily generated in Excel.

Edward Tufte offers insights on interface design

I don’t have specific suggestions for how to develop better ones beyond asking prospective customers how they represent the problem: what metaphors, paper aids, and mental models are useful in managing the problem they might hire your application to solve.  One author worth reading this is Edward Tufte: he has written four beautiful and insightful books on presenting information:

He periodically offers a one day course that presents material from all four of these books.

A New Product Can Succeed With Better Functionality
…And  A Worse User Interface

Be careful of relying on user experience / user interface as a proxy for “value.” Obviously a new entrant has to provide more value for a class of customer but many products offer substantially improved functionality or better cost but a user interface (UI) that is worse than the status quo.

  • Apple vs. Plato terminal was no contest for usability, but you could own an Apple.
  • If you compare Cisco routers to Appletalk user interface you realize why they had to create networking academies to teach people how to use them. Cisco ultimately had to use a different brand (Linksys) with a different paradigm to penetrate the home and small office / home office market.
  • Google Docs has a much less powerful set of features and less helpful UI compared to Word or WordPerfect except that it’s intrinsically shared on the web.
  • For an earlier example of the same thing look at wiki markup compared to publishing using Dreamweaver or FrontPage except that wikis were more easily shared by a team. But most product teams had great difficulty getting non-programmers to take part until a minimal WYSIWYG editor was added. This took years.
  • Early wikis didn’t worry about locking pages to prevent multiple overlapping versions to be saved over one another. When teams started using wikis to meet deadlines the benefits of a CMS became clear (and wikis added functionality to compensate).
  • If you want to take advantage of virtual servers you do not have anywhere near the visibility and management consoles that real datacenter managers have. But virtual machines are much cheaper for highly variable workloads and new features are now being added as virtualization adoption takes off.

Most new technologies have crappier UI compared to what their price/performance is obsoleting. It’s subsequent generations of product that attack usability / UI: even in the examples I cited the required original breakthroughs elsewhere:

  • tax prep software: encode tax laws–and keep them up to date–to calculate correct taxes/credits; interview format was a UX refinement that pre-supposed accurately encoding of tax laws.
  • treemaps: were a UX improvement targeted at representing large datasets
  • tornado chart: required that you could represent the decision model and solve convolution of all alternate futures accurately; early decision analysis tools didn’t have them but were still useful.

Capability -> Reliability -> Capacity / Cycle Time -> User Interface / Ease of Use

If you are attacking an existing category of solution then usability can be a key differentiator in the first product. But if you are trying to solve a customer problem in a new way this will often require interviews and careful observation to gain a deeper understanding of their needs and the constraints on your solution. The MVP’s you offer in this case may not be as “user friendly” provided they offer functionality or capabilities that are not available elsewhere. The challenge is that once you can reliably deliver results with the cycle time and capacity the customer needs you will have to address user interface issues or a competitor will go to school on your solution and provide one with equivalent functionality that is easier to use.

Related Blog Posts

1 thought on “Better Interfaces: Interview, Treemap, Tornado Chart”

  1. Sean, I’ve wished for an interview-style e-mail composer. (To begin an e-mail thread, not for replies.) Make it more like filling out a form, with a checklist to make sure I didn’t forget anything. Ask some key questions whose answers won’t be in the final message, but just get me to think better about communication — who is my audience, what’s my purpose, is email is even the best medium, and so on.

    If it were a plugin for my existing email clients, especially Outlook, (that is, not a new email client) and made my communication more effective, I’d pay for that.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top