Gerry McGovern - Dirty Magnets
The title of the talk referred to the sort of ill-defined links on websites that serve to distract users from their tasks. For example, "Quick links" navigation is all too often a bucket for putting in stuff that is irrelevant to users but valued by stakeholders. Gerry made a funny observation that surely this implies the existence of "slow links".
His preferred method for auditing the content of an existing site is to deploy a survey (via SurveyMonkey) with a list of all the site content and ask participants to rate 5 items of content in order of importance by placing a number from 1 to 5 alongside the relevant item. A consistent pattern emerges with 25% of the results rating the same 5 items, the next 25% a further 10 items. In order to be statistically significant, the survey needs to have between 200-400 respondents. Running the survey with stakeholders and customers can be used to highlight the differences in internal and external attitudes to content.
Some other points worth mentioning:
Audience-based navigation - e.g. Small business, Home user from Dell example - doesn't work as there is too much potential for overlap.
Customer/Advertiser relationship has flipped - search query "advertises" need to suppliers.
Mike Atherton - Brand-driven design
Mike's talk was all about the strategic use of branding to differentiate your product in the marketplace. He works for a B2B company Huddle that competes with the likes of Sharepoint for cloud-based collaboration services.
He used a technology lifecycle diagram from Don Norman's The Invisible Computer to illustrate how over time features eventually cease to be a point of differentiation for products. Referencing Stephen Anderson, he showed that emotional design seeks to increase motivation to use product as opposed to mere usability which is about removing friction.
Using a quote from Al Ries, he defined brand as "the single idea you own in the mind of the customer", a memorable and useful definition that allows you to focus on a single point of difference which in turn allows you to look at ways to segment the category your product fits within and reframe the conversation with customers. Another useful diagrams, the customer based brand equity model, takes account of rational and emotional responses in the context of an engagement journey. Brands can use personality to distinguish themselves - a wonderful example of the "Nauga", a character created in the 1960's by legendary ad man George Lois, to give an otherwise functional brand some personality (remind you of anyone we know?).
Mike round off his talk with some practical examples of workshops that he ran at Huddle in order to establish traits and values of their own brand in comparison with others. Plot index cards with own brand and other brands on strong/weak/positive/negative axes and write postioning statement on the back of the cards e.g. "Brand X is always _____ but never _____."
Milan Guenther - The Focal Point of Brand Identity and Enterprise Architecture
MIlan's talk was an interesting looks at a strategic design approach for dealing with large, complex enterprises. This is something I have become quite interesting in myself after reading a lot about the limits of the traditional consultancy approach in effecting lasting organisational change.
He basically advocates that designer need to be push beyond the interface and look at also having input into the design of the technology and processes within a business that are critical to the success and sustainability of any new initiative.
He recommended that IAs familiarise themselves with ArchiMate, a modelling language/tool for enterprise architecture as this fosters appreciation of the wider context/contraints for any new product/service within an organisation. He also recommends looking at the viable systems model, a diagram that can be used to represent any autonomous system at a high level.
Coaching and modular design frameworks are specific examples of the types artefacts that consultants can embed within organisations and ensure greater longevity for their design efforts..
Corolla Weller - RITE: Testing and a Business Driver
Corolla had a good phrase to describe the kind of waterfall testing done after development is complete and prior to launch - Alibi testing. In contrast, Rapid Iterative Testing and Evaluation (RITE) allows testing to happen much earlier and repeatedly throughout a project, provide an improved product with stakeholder input. Lucy has written a more in-depth write up on the approach.
The process basically involves testing and iterating on subsequent days within a tightly defined schedule (6 days), with each iteration focusing on fixing the issues identified in the previous day's tests. As a simple point of differentiation, RITE focusses on problems solved rather than problems identified.
Observations are collected on post-its during test by all stakeholders present including representatives from the business. Post-its are colour-coded according to positive/negative observations. The screen from the test are printed out and the post-its attached post-its during evaluation after observation. It is essential to come to a decision at the end of each evaluation in order to brief the changes for the following day's iteration.
The prototyping method they mentioned for providing the right balance of fidelity and speed of iteration was using static screens with hotspots. The method could be applied to any fidelity of prototype however given the commitment required from stakeholders, my assumption was that higher fidelity is better for facilitating feedback and decision-making.
Stefan Bussolon - Mental models as a common ground for cross channel design
Stefano's presentation attempted to show how mental models can be aligned with MVC design patterns in software and therefore represent a good basis for responsive design across multiple platforms, model first being conceptually similar to mobile first. Although I am a big fan of mental models and found his talk interesting, there was a lack of real world application to his ideas that left me feeling like it was a bit half-baked.
Users have needs, user goals are determined by needs.
User goals have cognitive representation.
Designers also work to create representation
Model -> Model
View -> Representation
Control -> Interaction
Franco Papeschi - A different grid: multi-channel user experience, the African way
Franco is a Service Designer Tim Berners-Lee's Web Foundation who has spent a lot of time investigation innovation in African economies. His presentation was very interesting but focussed more an a series of interesting case studies rather than making any particular point. A lot of the examples were quite familiar already however there were a few ones I hadn't heard of before.
Esoku - market information system
A standout quote supported by the examples was that "innovation comes from underserved users with a lack of existing resources".
Martin Bellam - IA in the touch screen era
Finally, a presentation from Martin Bellam, formerly lead IA at The Guardian on how he has adapted his approach when it comes to designing for mobile devices.
His first point was they he now ruthlessly prioritise use cases for responsive design to ensure that the mobile context of use is foremost in his considerations. When it comes to building prototypes, responsive frameworks like Bootstrap are an IA's best friend and they allow you to create artefacts that you can touch and interact with them. His top tip to every UX person is to learnt to code, even just a little bit, as otherwise you are are like a car designer that never drives a car.
His experience of getting the fidelity right involves adding enough design so that won't be rejected outright by stakeholders but not so much that visual desingers have nothing to do. A slightly different context when working at The Guardian with a well-defined style guide perhaps but just about any company will have some elements of style that you can use I guess.
He gave an anecdote about how frustrating user feedback on touch screen apps can sometimes be. A lack of on screen affordance for swiping highlighted by a small number of user tests resulting in the needless inclusion of chevrons to an otherwise beautiful design. In certain cases, you should be willing to ignore some feedback from user testing and rely upon your own judgement for what is right.
Once you have designed an initial responsive layout that defines the hierarchy of content, sketches should suffice for further breakpoints (assuming you are working with a half-decent designer) so long as you remember that lower fidelity demands higher documentation e.g. explanation of what you intend in the design. Oh and don't forget to add "zing", a placeholder for any stylistic frippery so the visual designer can indulge themselves.