Highlights from BADCamp 2013: Saturday Sessions, part 1
One thought: there's no such thing as a perfect project. If you say a project went perfectly, then you haven't talked to everyone involved.
Educating clients: Introduce the idea of "internet-level" or "Drupal-level" problems. I can't both make your main site really mobile-friendly and give it a video background. That's an internet-level problem. As web developers we can do quite a bit, but we can't do everything.
Why users are irrational and what we can do about it
Kevin Oleary, Acquia
Some important biases, and ways to (help users) avoid them:
Confirmation bias
People tend to believe most what confirms what they already know.
Provide answers, not raw data. People are not good data parsers.
When reporting test results, provide anecdotes, not statistics. Stories are strong.
When running tests, observe users' experiences, not metrics.
Congruence bias
People tend to rely on direct testing of their hypothesis and neglect indirect testing or developing alternate hypotheses.
Have only one or two direct, clear paths to a given task.
Ensure your customer surveys have a broad base.
Don't just ask people to evaluate your possible designs; run tests on other similar sites as well, especially direct competitors' sites.
Curse of knowledge
Experts find it hard to see perspective of novices.
Follow ubiquitous patterns where possible.
Capture video of actual site users. How are people actually experiencing your UI?
Test even design patterns you consider established. Your users may not be aware of the same conventions.
Illusion of correlation
Misinterpreting sequential events as cause and effect.
Limit feedback to only things related to the task. Limit possible actions in a given context.
When asking how a site update or redesign affected conversions, be fully aware of the context. What else happened at the same time? Was there a shift in the economy? In your competition? ...
Limit the variables when testing changes in a design.
Loss aversion
People prefer avoiding losses to acquiring gains, and remember losses much longer than gains.
This can increase customer retention if users can create something before having to decide to commit. Used ethically, it can drive stickiness.
Compare only future cost. Past investment in designing something has no value if it will cost more to fix its problems than to rebuild from scratch.
Maintain emotional detachment from the work you produce.
http://en.wikipedia.org/wiki/List_of_cognitive_biases
The Agile Drupalist
Adrian Jones, ImageX Media
Individuals and interactions over processes and tools.
Working software over comprehensive documentation.
Customer collaboration over contract negotiation.
Responding to change over following a plan.
Research shows greater chance of success. (Chaos manifesto, Standish Group)
Easier, less costly, to adapt to change.
Often better at managing risk.
Necessitates more involvement from stakeholders.
Better chance of building something useful.
Quicker feedback from end users.
Things to consider
Fixed budget/scope? May to take a more waterfall approach.
Uncertainty or end goal not completely clear? Agile will likely give better result.