LeanUX advocate with a passion for meaningful conversations
"Oh yeah,
so, what do they eat?
what do they read?
what do they drive?
why?
what do they do when they get up in the morning?
so, you're building a software product, great... what do they have open on their screens other that your product?
what else is on their desks?"
- the beginnings of deep thinking about customers, courtesy of SG Blank
@leemcivor champions integrity in research.
He advocates rigour, making sure you do something useful with those personas beyond creating them. Your research deliverables are not an end in themselves: they're not a measure of your success; your interpretations are. Also, make sure you attain quality in your findings by using as many participants as you think are needed for testing.
Poppy James from @bunnyfootsays says we should be maximising confidence in the data we gather; making sure it's representative. We rely on freewill and motive for participants so we should mitigate risk for volunteers. We should be transparent about the sample. Recruitment criteria must be based on good customer research e.g. what are the travel needs for a visit to the capital for non-londoners. When we look at systems at A&E - do we covet feedback from tuesday morning pensioners and/or the Saturday night reveller? and are we able to illicit feedback in a given situation? Consider the geographic and demographic in order to gain quality. Overcome as many pitfalls as possible and be transparent about those you can't.
@clivegrinyer talks about designing for experience.
Designing on hunches he quips: "I think it's called creativity". A faceless internet systems company like Cisco facilitate experience e.g. global telepresence connections between schools. People are key to experience, great stories can encourage tech companies to create great and appropriate solutions. We should look at the whole experience e.g. from on online purchase via Amazon - tracked logistics - to it not going through our victorian letterbox - to having to queue for our package at the sorting office. Wholistic visualisation of an entire experience enables us to design for its components. We should design with a view to see our future selves with the product. Strategy and decisions should be affected by deep thinking on people and their experiences.
@richardcaddick, from CX Partners talked about stealth research.
How can we observe chaos to mould our design of systems? Just watch people; listening at the sales desk. Annoyances and pain points arise from the language used. Listening in at call centres: people have very specific pain points when you do this. Things you can't guess at arise: men phoning a travel agent just wanted to know if there were places to watch the football - if they couldn't (find out) they wouldn't book. Forum conversations are also more genuine; notes that Google now has Discussion filter.
Stephena Broadbent considers Attention & Communication.
Giving attention to me the speaker, gives status. We have a wide array of communication channels. Some have less than 3 main contacts such as the telephone and Skype. Social networking, a broadcast medium, has many contacts. Voice requires immediate attention; broadcasting demands low attention. So we have an inverse relation between no. of contacts and channels of communication. We therefore have a suggestion that successful development of ideas should lean towards low attention media.
Jaimes Nel @gnva, from @liveworkstudio suggests that a linear approach to ideas are not entirely satisfactory.
He says that artefacts and ideas are the same thing. Taking a timeline for a concept through the past into today, may lay a path for a future trajectory. His talk reminded me of @blaurel's talk at #ixd11 on mapping emergent technologies.
@maccymacx took us through the methodical and rigorous process of observational data gathering and pattern processing (with software) in order to provide, proven and consistent persona creation based on real evidence.
Mary Cook at @uscreates advocates research via creating/designing better environments for research participants for improved engagement e.g. pop-up 'rant' boxes for teenagers, free coffee vans for migrant workers. Designing environments makes for better communication and ultimately involvement e.g women participating in breast screening interviews are still contributing to the project four months later.
Overheard a couple guys on the train the other day talking about photos, and how they used to shoot them and how it is nowadays. "Thing is" one says, "I used to take a lot of care over each shot, but now I just don't bother". I had a few thoughts about this UX wise. Convenience and low cost has made us less inclined to look was one; but maybe he's just older and not as inclined towards photography as he once was. The film process in some ways forced us to attempt to optimise our use and therefore to care. Unrolling and fitting a new film required care; as did extracting them. I definitely enjoyed my time mastering the Pentax K1000, but more than that; I can remember how it felt to get the best from it. Now, like a lot of us (and the two guys), I shoot on a few digital devices including a Nikon DSLR but feel engulfed rather than engaged. Quantity over quality I guess; and low cost equates to throwaway. I know there are some who take their digital gear and their photography seriously but for most of us, we've probably lost more than we've gained.
Reading through the lean principles, one of the things that struck me as vital is right there at the top of the list; Eliminate waste: eliminate everything that does not add value.
Nobody really wants to design by committee: clarity suffers, hierarchy reigns and free-for-all subjectivity glosses over customer focus and elongates each iteration. Persona development can alleviate that; the picture it paints is simple, digestible and memorable.Based on real user findings; collaborative persona building not only eliminates waste, it enables improved skills within the team around really understanding customers. In subsequent meetings, any new or developing ideas raised can be quickly assessed against the collaborative and patterned view.
No waste and no hard feelings.
So, we've got this horrid plain text list which needs tagging up into links.
1–5, 1–14, 5–3, 5–11, 5–34, 5–37, 5–44, 11–2, 11–16, 11–17, 11–18, 11–19, 11–20, 11–21, 11–24, 11–25, 11–27, 11–29, 11–30, 11–32, 11–34, 26–11, 26–12, 26–13
With the help of the regular expression find and replace built into Textmate we can sort this out in a jiffy.
Find: (\d+)–(\d+)
Replace:
$1-$2
Helps us out with the following output:
1-5, 1-14, 5-3, 5-11, 5-34, 5-37, 5-44, 11-2, 11-16, 11-17, 11-18, 11-19, 11-20, 11-21, 11-24, 11-25, 11-27, 11-29, 11-30, 11-32, 11-34, 26-11, 26-12, 26-13
Neat huh?
We used the \d expression to find numbers, but we could have used . to find any character. + means one or more. The brackets denote each variable which we can then use chronologically i.e. the first bracketed set can be used in our replacement string as $1, and the second set as $2.
More at: http://manual.macromates.com/en/regular_expressions
Easy peasy to do and may also be of use when Twitter roll out their next 'awesome' feature.
1. Install the Stylish addon on top of Firefox https://addons.mozilla.org/en-US/firefox/addon/2108/
2. Restart the 'fox and you'll see an S icon in the bottom right of the screen
3. Click the S icon and Select: Manage Styles
4. On the bottom right of that screen you'll see: Write New Style
5. Push the Write New Style button
6. Give your new style a name and add the following to the big central text box:
#recommended_users {
display:none
}
7. Click Save and watch those smug avatars vanish
You do? okay then. You know that to make your system better, you need to tap that thought.
They go about the system in ways you wouldn't, they know the ins, outs, feelings and nuances. There are small problems and irritations, tiny frustrations and maybe a big problem or two. You're going to tap that right? or you already do; you know how to ask those open questions that allow people to talk about their day and how they dip into your product. You know how to elicit those passionate feelings. You know how to find out what they make of it all and discover the pain points. You then know how to take that away, think about what it really means and how to interpret those thoughts.
So then you've really understood, shown empathy and your next iteration will rock, and I salute you.
Consider the flow you can attain when juggling several tasks on several websites and you arrive at Paypal for some reason. Hopefully it's to withdraw to your bank account, but first you have to login (our primary task). The highlighted button is to Sign Up, the Login button is below it in a muted gray. Now I always have to adjust my thoughts to make sure I go for Login. I'm drawn to Sign Up subconciously because it's orange hue is so enticing. My experience here is that I need to be wary when visiting Paypal so I click the right thing. The experience is switched when I do get inside - primary functions such as Submit are now Orange. The lack of consistency with web conventions and within itself need ironing out.
I always like to spot real life example as I go about my day. One that doesn't feel right is good service announcements on the London Underground. You are interrupted to be told "Here is a service update - all lines are running a good service". As most announcements are for delays or disruption most people make an effort to listen. If all is good then it should just be. You are running a service, I shouldn't have to be reminded you managed to get it right - I know; a waiter doesn't say "did you see the way I poured that wine, flawless, not a drip".
Once you're up with the prerequisites: http://docs.heroku.com/quickstart
If you have some specific gem dependencies create a new file called .gems in the root directory of your app e.g.
mislav-will_paginate --version '>= 2.3.11' --source gems.github.com
authlogic --version '>= 2.1.2' --source gems.github.com
then:
cd myapp
git init
git add .
git commit -m "commit to git"
heroku create
git push heroku master
heroku db:push
(if you don't get a running app after the db:push then you may have a schema discrepancy - like when I did by naughtily deleting columns within mysql itself)
REFS:
Sometimes you gotta wrestle some view into shape and the database might be part of the problem. Anyway, you end up in googleville or rails forums and don your hunting hat.
I needed to check for nulls, nils and plain old blanks to format my output nicely so I thought I'd end up with something like this:
<% case @person.address_2 when !nil, !'' %><%=h @person.address_2 %>,<% end %>
which doesn't throw an error, but it also doesn't really output properly, what does is:
<% if !@person.address_2.blank? %><%=h @person.address_2 %>,<% end %>
This tasty snippet allowed me to check for not *empty* before outputting an address section with a lovely comma (I didn't want a nasty looking string of commas when I didn't have an address).
As ever, they'll be a best practice ruby way but this worked for me, for now.
