Just a very quick post to let you that I recently created a Ning community to allow designers and user experience people who are working in (or interested in working in) Open Source and Free Software communities to share their experiences, their projects, their questions and their mental health breakdowns!

If this sounds like something you might be interested in, come join us here: http://www.designintheopen.org/

I’ve been known to say in public places that designing with a community is nothing like design by committee. And, certainly that’s been my experience to date. Not to say that designing with a community is painless either! These days I feel as though I have a vast swathe of constantly raw virtual flesh put out for cutting. Checking email is a lot more scary now than ever before!

Recently we’ve been getting some feedback from within the Drupal community that our approach is not open enough. The designer in me finds this devastating – by ‘designer’ standard I think we’re already being profoundly open in terms of both the way that we’re working (our practice) and also what we’re designing (our output). But, by open source standards, I completely understand the feedback. Angie Byron (@webchick) recently left a comment on an earlier post that describes some of the issues that are emerging. I’ll quote a little from that now so you can get the picture:

On something like groups.drupal.org, everyone can be a content creator and make new posts which are equivalent to everyone else’s posts in “primaryness.” While we have tools like “sticky posts” to draw attention to particularly important things, everything else is open to everyone and has a real collaborative (if chaotic) vibe. This is more like “the Drupal community show, with special guests Mark and Leisa.”….

…. But yet, the current model feels to certain members of the community like “shout it out into the darkness and hope someone’s listening” collaboration paradigm, when they’re used to much more direct interaction like pinging webchick on IRC and saying “Hey! I’m upset about something. Let’s talk.”

I can’t help but preface my response by saying that I’m usually available for hours at a time on IRC to Drupalers and more than happy to be pinged with this exact message, although having said that, I’m no @webchick in the Drupal community so completely understand why this hasn’t happened as much as it might.

I do want to share back some of my thoughts on ‘equivalence’ and ‘the Mark & Leisa show’, as the way that this is playing out is not entirely accidental and there is a certain amount of ‘design’ to the way that we are communicating within and outside of the Drupal community (having said that, it is a constantly evolving design and you are probably being party to a big evolution right now!)

Top Down/Bottom Up

Another thing I’ve been known to say in public places is that design projects like D7UX and the Drupal.org redesign project aren’t the kind of projects that can be initiated from within a community, that is organically and ‘bottom up’. I think it’s true. Look at the efforts of the hardworking usability team within the Drupal community – there are some very smart people there and they are working terrifically hard and making a big effort to improve the users experience of Drupal – one issue at a time (using the issue queue infrastructure typical to opensource development). They have certainly taken big steps to improve Drupal’s usability, but (and of course, this is open to dispute) I believe there is a fairly profound change that needs to come about in order for Drupal to achieve what I understand it would like to achieve – that is to make it’s powerful tools more available to people with less or different or, ideally, even no web development skills.

I wouldn’t like to say it was *impossible* for this to come about from within a large community, but I think it is self-evident that it is highly unlikely. It requires a different type of methodology to what is found in open source communities at the moment for this kind of change to be designed, at least, within any kind of useful timeframe.

We need to find a way for a good design methodology to work in an open source environment. That’s one great big hairy challenge.

The Open Design Triangle

Back in my ‘agency’ days, we used to use a diagram of a triangle to try to explain to our clients some of the facts of life around features and cost and time and quality, it is quite like the one I’ve included above except the three corners said ‘Time/Quality/Cost’ and the idea was that the client could choose to move two of the corners but that one had to remain fixed. The triangle couldn’t get any bigger but it could change shape.

I’ve been thinking about this model recently in terms of the way we’re working on the Drupal project, and for open source designing in general, and I think a similar principle applies, except that the corners are now labeled

1. speed/time
2. quality and
3. inclusivity/openness’.

In *this* project time also equates to cost (as we are being paid to work on the D7UX project) but obviously this is not always the case. And as with the commercial model, you can move any two of these corners but not all three.

Time

Time is a massive issue for us in this project. I would estimate that at the moment we are spending at least 65% of our ‘paid’ time just ‘communicating’ on this project and just 35% actually designing. That’s scary for us – there is a lot to design and not a whole lot of time to do it. We would LOVE to be spending more time designing, but as I’ve said before, it’s fruitless and foolish to do so in isolation.

On top of this ‘paid’ time, both Mark and myself spend a lot of our own time on this project and in that time we are almost invariably blogging or in IRC or responding to email communicate we’re receiving on the project. I’d estimate for myself that I’m probably contributing an extra 20-25% of time beyond what I’m being paid to do on this project. That doesn’t make me any great hero, I know, but that’s the current landscape.

Quality/Integrity

I’ve written in the past that I believe that design is no place for democracy. Open design is amazing because you can have so much feedback and insight piped in throughout the process, and as I hope is evident, we are doing everything we can to encourage this engagement in our process from the outset (where it is arguably most important).

However, design decisions at a system level (like the ones we’re working on at the moment) shouldn’t be made issue by issue and by consensus – not if you want a great design, a great user experience. A good design comes from a strong, unified vision that is articulated by experienced designers. The power of this clear design vision is that, going forward when design decisions do move down into issue queues (which, over time, they will), we are all able to discuss design issues and make design decisions more articulately and more effectively because of the foundations that have been laid, through the ground work that we are doing right now.

‘The Mark & Leisa Show with Special Guests’

As I understand it, there are three main reasons why this impression is being created. Let’s look at them one at a time.

1. We’re constantly shouting about our work
   Especially in the early stages of this project we are spending a lot of time effectively saying ‘look at us, look what we’re doing’. Don’t worry, it feels pretty uncomfortable for us to be doing this, but if we want this project to succeed then we have to do it. The greatest risk to this project is that we don’t engage the Drupal community (engaging ‘outsiders’ is also very important but nowhere near the same level of risk), and that we don’t engage them early enough in the process when the big decisions are being made. There is no way that we can just send out one message and know that we’ve reached everyone. We have to shout, as loudly as we can and often. I don’t think we have a choice if we want this project to succeed.
2. We’re using video
   We chose to use video for a few reasons, partly because ‘show and tell’ is often easier for people to consume than text, and also partly because we want to come across as human beings with feelings (in the hope that people will consider this as they’re drafting their responses).
3. We don’t often get ‘conversational’ around the feedback we receive
   We are receiving feedback and insight from lots of different people in lots of different places – we’re doing this so that we can maximise the level of engagement and involvement that we get with the project. Very often we stay a little removed from the feedback that is coming in, for a couple of reasons.

• Firstly, we don’t want to get involved too early – we learned from our work on the Drupal.org project that if we stick our noses in too early we can skew the direction of the discussion, and we don’t want to do that. We love it when conversations evolve amongst the community and we watch very closely how those play out. We need to give a thread time to evolve to see what trends emerge.
• Secondly, time. See above.

Consensus Driven Design?

The way that we are currently engaging with the community is very different to the way the community currently gathers to discuss and resolve issues – which is very much consensus driven.

I cannot say often or loudly enough how much we crave communication with the Drupal community on this project, but in order for us to do our job well, we need to engage in a somewhat different way.

We can’t argue every single point at the moment that it is raised. Our process doesn’t work that way (we don’t know anywhere near all the answers at the moment, we’re still devising the strategy to make the questions that we’ll then set about answering, with the assistance of the community). Also, see ‘time’ above.

Having said that, I think that we are striving to work in a consensus driven way. We’re doing this by creating larger artifacts that we can gain consensus around. Things like our Experience Strategy, our Audience Matrix, our Design Principles for example, are ways that the community can gather around the work we are doing and we can get some kind of concensus about the best way to define our strategy.

In the recent release of our Initial Concepts, we specifically pulled out four artefacts for discussion, with the aim of gaining concensus around them before moving forward (being the header, the overlay approach, the inline editing and the ‘direct manipulation’ tool).

It may not look exactly like the way that concensus driven development works, but the principle still holds… at least, that’s the outcome we’re trying to achieve, within the contraints of time and quality (see above).

Where to now?

There is no right or wrong way to do this, yet. The work we’re doing on this project will, hopefully, be used as reference for future projects and I’m sure they’ll look at some of the things we’ve done and say – great! and others they’ll look at and say – rubbish! We’ll probably do this ourselves before the end is reached!

I’d  love to find a way to more effectively synthesis all the feedback we’re receiving and to share that in a way that everyone can consume more readily. Again, I’m not sure exactly how to go about that yet, but I am fairly sure it’s not to just talk in one place.

We are listening. I think we need better ways to show that we’re listening. I’m not sure what those are yet. I’m going to think on that some more and I hope you do too and let me know what you come up with.

A final note on Fear
I wanted to wrap by sharing another part of WebChick’s comment, which resonated deeply with me.

some feel like they don’t have expertise in this domain and are really intimidated to jump into the fray. They’re scared to say anything bad because they’re convinced that their opinions will get immediately shut down, and that they’ll offend you guys.

So much of this project and this way of working is about fear. I know I feel terrified by this project almost every day I work on it, for so many reasons. And fear doesn’t often create a great environment for communication. I need to think about this some more (and no doubt have a whole other blog post brewing), but I thought it was worth throwing that out for your consideration.

Thanks for taking the time to read all of this. Be brave. I look forward to hearing your thoughts.

It was pretty obvious from the outset that we’d need some design and UX skills to get us from one end of the Drupal.org redesign project to the other. It was less obvious how important our ‘social’ skills would be – and unsurprisingly, we learned a lot about good and bad ways to share the design process with a community along the way.

Here’s a few ‘social skills’ we learned:

1. you need to take responsibility for the way that your community behaves: it’s not in any way productive to associate the way that a community is responding to you by blaming the community or even the individuals in it. If you respond that way you’ll never be able to improve the situation. As with every relationship, the only person you can change is yourself. If you’re getting a bad vibe back, the first thing you should do is check your tone and content – what are you saying? how are you saying it? can YOU improve the way you’re communicating. The onus is on YOU to get it right.
2. tokenistic involvement is a waste of time: if you don’t really care what the community has to say on a subject, don’t ask them. If you do want their input, take the time to design a way for them to interact with you in a way that gets the best from them. Be creative, put a bit of thought into it. Avoid polls and and use surveys with care – you might feel as though you’re involving the community because you have ‘numbers’, but do you have real involvement. Ask yourself what the community knows that you can benefit from, then consider the best way to help them share that knowledge and experience with you.
3. ask for specific feedback: if you want to get good feedback from your community, tell them what you want feedback on. We *didn’t* do this much during the Drupal.org redesign – instead I was trying to keep it ‘neutral’ and not influence what and how people gave us feedback – we learned that by asking for specific direction we not only got excellent feedback on the issues we highlighted, but others as well. Without direction the discussion tended to be less helpful and was more likely to get personal (not in a good way!) This will also help you to get feedback on more than just the homepage.
4. give examples: if you want a particular kind of response from the community, it is important to provide an example for them to follow and really great instructions to participate. For example, when we were doing the ‘crowdsourced wireframing’ I included a picture of one of my not very elegant wireframes so that people had a sense that their submissions didn’t have to look ‘designed’. If there are instructions to participate, make sure these are as clear as possible. Then make them even clearer.
5. wait… wait… wait… engage! once you post something for feedback, go away and make a snack and do NOT get involved in the conversation immediately. This is probably the most difficult rule to follow and one that Mark and I had to coach each other on (and occasionally police! – step away from the computer!) throughout the project. If you dive in and start responding to the first few comments, what you unintentially do is skew and retard the conversation. Rather than exploring a broad range of issues and allowing key points to gradually evolve, the discussion focusses on whichever points you have responded to, everyone starts to focus on those few issues. The richness of the feedback is lost because you dive into detail too quickly. Rather, wait until at least a half dozen people have posted (or 12hrs has elapsed, whichever is soonest) and see what the trends are in the feedback, then start getting more involved in the conversation.
6. admit errors quickly: the only exception to the rule above is if you’ve stuffed up. In this circumstance you should admit the mistake quickly so that the conversation doesn’t focus on your error. In one iteration of our redesign we accidentally omitted a very important call to action (I know… how could we?!) As you can imagine, that oversight dominated the feedback we received and by the time we responded (way too late!) things were getting a little frenzied. We should have been keeping a closer eye on the situation and stepped in as soon as we realised our mistake.
7. don’t go dark, but don’t respond to everything: there is a balance in the correct volume of response that you need to aim for. It is really important that you don’t disappear (even if you get really busy) – the community needs to know that you are there and that you are listening. On the other hand, don’t feel as though you need to respond to every comment that is posted – unless you are only getting a handful of responses. As a rule, aim to respond to trends and issues not individual comments. Feel free to occasionally respond with a simple ‘I’m here and listening, I don’t have the answer yet’.
8. lead by example: it’s an oldy but a goody - interact with the community in the way that you would like them to interact with you. Be polite and respectful. Others rudeness or bad behaviour is no excuse for you to let loose. It’s up to you to set and maintain the standard of communication you want the community to engage in.
9. assume good faith: it’s a general rule of interacting with others online, but keep it at front of mind especially when you’re putting your own work out there for review and, therefore, more likely to feel a little defensive. Text is a tricky medium for communication – people might sound like they’re being mean or overly critical or agressive when they’re just not great at communicating (or you’re feeling defensive and read everything as an attack!), or being a little lazy with their words, or created unintentional meaning. Always assume that people are trying to be friendly and constructive and helpful if there is any room for doubt at all. In fact, even when it is evident that they *are* being a little mean, it is often useful to deploy this rule – play dumb and be extra nice. Don’t waste time fighting or being a smart ass, or just being mean, or engaging with others who are. Focus on the task at hand – doing good design.
10. be a human: I think this is the absolute most important thing – don’t assume a Voice of God, don’t pretend to be infallible or to know everything. Don’t feel as though you have to use very big words all the time. Swear occasionally (if your community is ok with that). Admit that you are nervous (or outright terrified, if that’s the case). All of this is allowed and encouraged. Communities are made up of people, of human beings and you are but one of them. Use your real voice and speak honestly. Be open.

Continuing in the series of ‘Things we learned from the Drupal.org project‘, this post actually starts off in the comments of the last post (design by committee vs design by community) where Keith picked me up on the statement that design should never be democratic and asked ‘Could it be? Or at least closer? And how to do that?’

Ah, democracy. It is a beautiful theory, but only – as with so many things – when applied correctly. Democracy may be great for deciding on a government, it may be great for Pop Idol (hrm..?) there may be other places it is great and noble… but design just isn’t one of those places.

When designing with a community what you should be aiming for is participation not democracy. Make your design process as open as possible, but don’t be fooled into thinking that because people ‘voted’ for a particular design, that is is the best decision, or even a good one. It’s probably not.

There are two key reasons that I believe this to be true:

1. Good user experience is hard to evaluate when not in use – when you give people a screengrab or even a prototype to evaluate, people will tend look at things from a visual design perspective (highly subjective), and often a ‘heuristic’ perspective (usability conventions, best practice, what ‘users’ do and like etc.).These perspectives are valid and interesting, up to a point – but they come nowhere near being as valuable as the observation of a designer, or actually observing someone performing tasks that they would do every day using your design and seeing how it works for them. I’d give that trumps over popular opinion any day.By putting a design out there and asking people for their feedback, you are actually giving them a really difficult task. It’s hard enough for those of us who do it professionally (and there’s plenty of research to show that our opinions can vary wildly) – it’s not really fair to expect your community to be able to make a good decision about whether or not a design will work well based on just taking a look or clicking through a prototype.
2. Your community are domain experts, not design experts – the best thing your community can do for you is tell you what you need to know in order to design well for them. Most of the time, they are not designers. They don’t have design training. Why are we asking them to do design work?If I could find my copy of Bill Buxton’s Sketching User Experience (which I have conveniently misplaced on the day of London UX Bookclub, d’oh!) I’d find the part in it where he talks about how ‘reading’ design, interpreting sketches, is actually as much of a design skill as doing the design in the first place – it’s just one we don’t talk about and don’t place any value on. Part of the reason designers often snort at the feedback given to them by clients (or community members) is because of a lack of design literacy in their feedback. Well, of course. They’re not designers.

Your mission when designing with community is to facilitate the community to make good design decisions by working with the information and insight they provide to give them good design and help them understand the design strategy and how and why it works.

Giving the community a true and meaningful voice in the design process is so much more empowering and respectful of them than letting them vote for which design they like the best. Letting a community choose a design by popular vote is, in my opinion, relinquishing your responsibilities as designer.