Think Info

Exploring the information space

Usable Usability (EuroIA 2013)

Kicking off EuroIA in Edinburgh, I attended Eric Reiss‘ workshop – really, as he admitted, a master class – on Usable Usability. The theme echoed his book of the same name, released last year.


Eric didn’t hold back with his class; it was heavy with information, ideas and practical approaches to improve usability; not by focusing on any specific, but largely by poking fun at the endless stream of counter-examples.

In all, usability is about people interacting with things. While it is easy to claim that usability stops at the point of “does it make the action achievable?” (whatever it is that is supposed to be done with it), the reality is that there is far more to something being considered usable. It is about the ease with which it can be used.

Put simply, usability is an equation:

Ease of use = Functional + Responsive + Ergonomic + Convenient + Foolproof

  • Does it accomplish what it “says on the box”?
  • Does it provide feedback that it is doing its part?
  • Does it provide an interface to do, with minimal friction and discomfort?
  • Does it provide an interface that is simple, logical, streamlined and self-contained?
  • Does it cope with the edge cases of its use?

This question of usability can also be approached from a second angle, where many of the elements cross over to a large extent with the first, but allow the mind to focus on other layers of detail:

Elegance and clarity = Visible + Understandable + Logical + Consistent + Predictable

  • Visibility is about the honesty of communication – that links and references really pertain to their subject
  • Understandability is about language, and a shared frame of reference
  • Logical relates to structure and continuity
  • Consistent deals with putting in the effort to ensure that any two parts of your product/communication are clearly associated, because they speak the same verbal/visual language
  • Predictable is about playing off generally-accepted themes and conventions, whether global, regional, or industry-specific as relevant to the audience/subject

A key technique for ascertaining usability was identified as the elevator test: when the doors open from a lift, how clear is it what your route options are, and how easily can you determine where you want to go to complete whatever task it is you took the list for? Apply that same consideration to the interfaces to your communications/product; make them pass the test.

Finally, two sure signs that there is something lacking in the usability of a design:

  • It has additional instructions added to it – a sign stuck to it – explaining what isn’t clear by default
  • People ask “why?” about it, and how it works.

Raw notes

  • Usability is situational. It also changes over time.
  • Is your trash can too small, or is it not being emptied enough? It’s a service problem (not usability)
  • Usability: It does what I want it to do. It does what I expect it to do.
  • The average user – we’ve all heard about him – is kept in a vault in Switzerland.
  • The visual hints provided in an interface reduce the need to think about the value/function of the element.
  • Ease of use = Functional + Responsive + Ergonomic + Convenient + Foolproof
  • The function of usability in the digital world maps very specifically to the real world.
  • The lack of skeuomorphism makes an interface hard to understand
  • Forms are your only point of true interaction with the end user in digital – they are the most important interface.
  • Forms are the point that must be functional
  • If your content people don’t talk to your developers… who knows why people won’t enter content in the correct format.
  • In the EU, it is illegal to ask for information that is not directly relevant to the transaction.
  • Things only work for the people who have programmed them – in their one perfect scenario
  • Functional things to consider
    1. What is the goal of the stuff
    2. Do forms work
    3. Can forms be interrupted
    4. Are there edge cases outside the norm
    5. Are there rigid input methods
    6. Are there other ways to complete
    7. Does check-out work?
    8. Will it become less functional over time
    9. Will it work on all browsers, platforms, devices?
    10. Can you optimise your graphics for faster loads?
  • Responsive: things change to give you feedback based on your actions
  • The timer icon has a huge amount of information packed into a very small image
  • The myth of two-way communication: it is a sequence of one-way communications. Action + Acknowledgement + New action
  • Most responsive/adaptive design is about responding to the devise. Screw the device. Respond to the needs of the user.
  • We have all the hardware built into our devices to know all about its context.
  • Content must be contextual to its environment – so it makes sense to the person using it.
  • Responsiveness needs to work as well offline as online. And these responses need to be in harmony.
  • 10 responsive considerations
    1. Does a button respond to click?
    2. When a file is saved, can you see the change?
    3. Does the cursor react when something is interactive
    4. Does design respond to the device
    5. Does the content respond to the environment
    6. Is there ongoing feedback during long processes
    7. Is a physical object also giving useful feedback
    8. Is feedback timely
    9. Do you understand icons and response mechanisms
    10. Do you wish your device would say what it is thinking?
  • Many sites depend on mousploration to figure them out
  • 12 principles of ergonomics
    1. Work in neutral postures
    2. Reduce excessive force
    3. Keep everything in easy reach
    4. Work at the proper height
    5. Reduce excessive motion
    6. Minimise fatigue and static load
    7. Minimize pressure points
    8. Provide clearance
    9. Move, exercise and stretch
    10. Maintain a comfortable environment
    11. Enhance clarity and understanding
    12. Improve work organisation
  • There are people – our customers; processes – things that need to be done; and technologies – the mechanism.
  • The problem with the ipod shuffle – you can’t clip the damn thing on anything without reshuffling everything
  • Radio City Music Hall includes columns – not for structural support, but to facilitate crowd flow.
  • The updated version of Fitt’s Law: big buttons are easier to hit than small ones.
  • Fear, uncertainty and doubt will break usability faster than anything else.
  • 10 ergonomics to consider
    1. Are the buttons big enough?
    2. What about when using your finger as a cursor
    3. Can you catch drop-down menus? Or are there timing issues?
    4. Are you providing keyboard shortcuts
    5. Are you providing tabbing?
    6. Are elements required simultaneously visible simultaneously?
    7. Are screen elements getting in each other’s way?
    8. Can you design a “silent usher” to help users?
    9. Are visual gimmicks making your stuff difficult to use?
    10. Are there illogical sequences or workflow interruptions?
  • Convenience – are buttons where they are needed most?
  • Information architecture is whatever you do.
  • What is information architecture? Look at a fruit stand. What sells most? With variety? It’s up front.
  • Your add on a cardboard box is bad usability. Give me a storage categorisation matrix instead.
  • Hotel: has mirror, hairdryer, plug. Not physically possible to see self in mirror while holding drier. Usable?
  • User experience is about choreographing interactions – encouraging optimum paths.
  • Any time anyone says “for your convenience…” you know you are about to get screwed
  • 10 things to consider for convenience
    1. Is everything needed for task completion available?
    2. Group related content better
    3. Can you use colour or other visual signals to differentiate areas or concepts
    4. Do multimodal processes get in each other’s way?
    5. Do you really understand the user’s need
    6. Can you educe swaps between on- and offline functions?
    7. Have you given users reasons to love your stuff?
    8. Are you providing useful content? If not, what is missing?
    9. Do you spin inconvenience?
    10. Can you eliminate duplicate information capture?
  • There is always a bad reaction to change, even when it is an improvement.
  • It’s when you are outside your comfort zone that you notice that things are falling apart
  • There is not necessarily anything wrong with redundant information. Put it where people need to read it.
  • Can’t remember how to use “it”? Chances are you won’t bother using it.
  • Customisation is what I do to the page. Personalisation is what the system does to optimise for the user based on context.
  • Undue diligence: what happens when lawyers stick their fingers into interface design.
  • 10 foolproofing techniques
    1. Can you give people multiple ways to respond?
    2. Are your error messages understandable
    3. Can you speed up response times and refresh rates
    4. Are there physical deterrents?
    5. Are your messages confuses with system warnings?
    6. Are you getting in the way by being overly helpful
    7. Are personalisation tools remembering now-irrelevant behaviour?
    8. Are there cognitive clues and guideposts as people interact
    9. Are your instructions kept to a minimum
    10. Is the foolproof system worse than the problem?
  • Elegance and clarity = Visible + Understandable + Logical + Consistent + Predictable
  • The elevator test: the doors open, what can you do now? Where is what you are looking for?
  • Any time you see a sign, there’s a usability issue.
  • If people don’t see a control, it does not exist.
  • Cut down the noise. But there is a point where you can take away too much information.
  • When using nesting to associate, if you don’t have content for all your nests, you can’t communicate
  • Banners at the bottom of an interesting page get more attention than those at the top (banner blindness)
  • If you have critically important info, putting it in a “banner location,” it will be overlooked
  • Don’t attempt to always be above the fold; embrace the fold. Make it obvious. People will scroll.
  • Links need to be blue and underlined? NO. Links need to be distinct and recognisable.
  • Above the fold:
    • Branding and nav
    • Help/contact
    • Internal search
    • Shopping cart
    • Language change
    • Key input and output
  • 10 visibility techniques
    1. Do you suggest information is available when it really isn’t? Or hidden away somewhere dumb?
    2. Is something physically blocking the view?
    3. Does important info look like ads?
    4. Do you forget important info?
    5. Does the fold separate relevant info?
    6. Is it scroll-friendly?
    7. Is there a paywall getting in the way?
    8. Does every page meet the elevator test?
    9. Are you designing to reduce clutter, or just to make stuff prettier?
    10. Are you using internal/proprietary terms instead of ordinary language?
  • Do your links let users do what they promise? Or are they a gateway to a paywall?
  • When talking information, the information not provided is usually more important than that provided.
  • Answer the questions people will have, but anticipate the questions they didn’t think to ask
  • Most information is provided via shared frame of reference. (Make sure it’s relevant.)
  • Icons are never fully understandable. (That’s why every phone icon has words underneath.)
  • 10 ways to make stuff understandable
    1. Are written descriptions accurate and comprehensive?
    2. Are you using abbreviations of difficult words?
    3. Can you see touchpoints across channels?
    4. Are images building proper shared references?
    5. If you list prices, do people know what they include? (tax, shipping)
    6. Are there pages or processes that don’t make sense outside the local geography?
    7. Do you rely on icons too much?
    8. Are there physical constraints (small text box) that prevent building the proper reference
    9. Are you making comparisons? Understandable?
    10. Are your written descriptions misleading?
  • Types of logic: deductive reasoning, inductive reasoning, retroductive inference
  • Retroductive inference: the reuse of skills and being able to guess usage from past experience, despite small variation.
  • If you hear yourself asking “why?” there is a usability problem.
  • The disconnect between your expectation and how something is designed to work is a usability issue.
  • 10 ways to make stuff logical
    1. Do you find yourself asking “why did they do that?”
    2. Are there responsiveness issues that make you wonder why something is/isn’t happening?
    3. Are there ergonomic problems resulting from design?
    4. Are there situations where you think: “can we make this easier?”
    5. Did you take an action that disrupted a process? Why?
    6. Do things do what they look like they should be doing?
    7. Do your optimal use cases map out easily? If not, there’s likely a problem
    8. Does the flow make sense? Is it clearly moving towards the end user’s goal
    9. Are your shared references strong/clear?
    10. Does the (browser) back button break the process?
  • Use a consistent language. Different word? It means something different.
  • If things look the same, they should behave the same. If they look different… duh!
  • 10 steps to make things consistent
    1. Do design elements that look the same behave the same? They should. (consider from both sides)
    2. Are you overlooking “design best practice” (shared cognitive framework) just to be clever?
    3. Is consistency playing second fiddle to creativity?
    4. Do you have multi-function buttons of knobs? Less buttons is not always better
    5. Are you reusing icons for different functions? Redesign or drop.
    6. Can you increase functional uniformity across your product/service?
    7. Does your stuff require prior knowledge to be understood?
    8. Can you use colour/visual grouping to show relationships?
    9. Has your stuff been “patched” to cover up basic consistency flaws?
  • Don’t make something so foolproof that it gets in the way of the superusers.
  • Patches generally indicate that there is a basic problem that needs to be solved.
  • Warn of invisible conditions: Hot? Cold? Bright? Loud?
  • 10 ways to make things predictable
    1. Does it draw on past experience? If not, cognitive triggers?
    2. Are there knowledge prerequisites? Can you provide them unobtrusively?
    3. Have you clearly expressed your expectations?
    4. Are the steps in a process enumerated clearly in advance?
    5. Are you trying to get people to achieve the same thing they are trying to solve?
    6. Do you understand design patterns and best practice that match what you are trying to accomplish?
    7. Are you providing clear warnings where there is danger?
    8. Do your social media tools align with your guidelines for other content?
    9. Instructions… too many? Too few?
    10. Is your design taking people by surprise (does it fail to make immediate sense)?
  • Don’t work based on user centred design – go with user driven design. Get real users to help you design.
  • “Lamb with peas” cat food… sounds delicious… No one asked the cat. (They can’t eat vegetables)
  • To get usability budget, make the cheapest important change anyway, then show the boss the money.
  • Inventions are not innovations. Inventions can be accidents. Innovation is abstract association.
  • Invention spawn innovation which becomes best practice. If no new innovation, it becomes habit.
  • The purpose of innovation is to solve a problem. If you don’t solve a problem, you will create one
  • There are many contributing factors to the Titanic sinking. Change any one, and likely history changes too.
  • You cannot take an isolated incident and apply it as the standard across the board.

Comments are closed.

%d bloggers like this: