How do you make design decisions in a team with a flat hierarchy?

May 23, 2012 8:31 AM

How do you make design decisions in a team with a flat hierarchy?

I work for a small NGO. We are a very small team with a very flat hierarchy. Decisions are almost always made by consensus and with the participation of almost all employees. Currently we are revising our "corporate design" together with a graphic designer. However, the members of our team have all very different ideas about what the future design should look like. And it seems like we could discuss about this issue forever because it is a lot about personal taste. In corporations there is probably a CEO who makes these kind of decisions. We don't have anything like this and don't want to introduce one.

I can imagine there are similiar problems in other corporations / organizations. I would like to know how you make design decisions at your work place. Do you have any best practice advice?

posted by jfricke to work & money (21 answers total)

Here's Valve's Employee Handbook

They're a billion dollar corporation with a fairly flat heirarchy, and built around self-organization.
posted by empath at 8:32 AM on May 23, 2012 [4 favorites]

How many people in the organization as a whole? How many in your immediate group (i.e. how many people are doing the same job you do)?
posted by phrontist at 8:34 AM on May 23, 2012

Everyone puts in a design concept/idea/parameter/whatever. That gives you X picks. Everyone ranks everyone else's picks first, second, third, down to X-1. Give each concept X-1 points for each first-place vote, X-2 for each second-place vote, and so on.

Once everyone has voted, the highest one wins, or eliminate the lowest-scoring pick and revote, then do it again until you're down to one. If two tie, they're both out. If the last two tie, you give the graphic designer the tiebreaker.
posted by Etrigan at 8:38 AM on May 23, 2012 [6 favorites]

15 people. Nobody does the same job as me but at least six people are involved in design decisions.
posted by jfricke at 8:39 AM on May 23, 2012

You can always just pick someone to be in charge of that project and let them make the call. That doesn't make them CEO, it's just a relatively arbitrary choice that saves everyone's time.
posted by empath at 8:39 AM on May 23, 2012 [1 favorite]

Seconding @empath. Valve's employee handbook is solid.
posted by nickrussell at 8:46 AM on May 23, 2012

Check your corporate bylaws, assuming you have any. There may actually be a decision-making process in there that you've been ignoring.
posted by valkyryn at 8:51 AM on May 23, 2012

A structured approach to what empath is describing is called a RACI. You document the types of decisions that you have to make, grouped by categories. So it could be "Brand Decision", "Style Decision", "Technical Decision", etc. with the specifics listed for each. Then you list all of the different people, taking into account their roles and responsibilities.

You make a spreadsheet with all of the decision types in the first column, then have rows for each of the people. Then you assign either an R, A, C, I, or for each of them.

R means responsible; they are the person who must get this decision made
A means accountable; they are the person who is accountable for any decision blow back.
C means that they should be consulted in the decision process for their input, but are not responsible nor accountable
I means that they should be informed of the decision
means that they aren't involved in the output.


You can develop the RACI with all people present. The key is for people to self identify and for others to understand who is involved.
posted by dobie at 8:53 AM on May 23, 2012 [2 favorites]

sorry there should be a [blank] in there after the I and as the last item. I used html brackets by mistake.
posted by dobie at 8:59 AM on May 23, 2012

I would discuss this issue with the designer, they should be able to help your organization work through the process.
posted by jade east at 9:13 AM on May 23, 2012 [1 favorite]

Thirty-five. It's a corporate decision making game.

You might want to run a test with a less exciting question first so you get how it works and you'll probably need a group of 8 or so people at minimum.
posted by chairface at 9:39 AM on May 23, 2012

Good design is about way more than personal taste. See if you can steer some of the conversations towards questions like "how is this design element helping to accomplish our goal?" or "does this design element express our mission?" That often gets people thinking about the design in more concrete terms than just "I like blue" and "I like brown." Even colors, which would seem to be a completely subjective area, can be discussed in terms of how they effect the feel and personality of the work - should it be calm? sturdy? welcoming? cheerful? somber? (This also works for font choices.)

And if you do end up getting right down to half the group preferring light blue to dark blue or whatever - an even split - remember that you hired this designer for his or her skill and insight. He or she can serve as a tie-breaker (appealing to authority, if you will) if you get to that unfortunate point.
posted by agentmitten at 9:40 AM on May 23, 2012 [1 favorite]

i'm a graphic designer, and as one, i wouldn't want to be engaged until your organization knew what you wanted to do with your rebranding/redesign, and came to me with a brief that outlined your goals. it would then be my job to visually interpret your brief. it wouldn't be my job to help you decide what your goals are with the redesign, and working with a team who isn't unified and all have very different ideas of your brand vision is a nightmare to deal with. presumably you hired a designer based on their skills and their visual aesthetic. if they are good, their initial options for you should be grounded in how their design interprets your brief/goals and they should be able to defend and explain how each design accomplishes that, from the typography to the color to the layout, and the actual graphic itself. the design that works the best in accomplishing your goals as a brand is the successful one, regardless of personal taste, and that is the direction you take. sometimes it's a matter of combining more than one of the designer's initial ideas to get to that point.
posted by violetk at 9:47 AM on May 23, 2012 [1 favorite]

the members of our team have all very different ideas about what the future design should look like

Yeah, this is very common.

The thing is, you shouldn't be trying to do your designer's job. Tell the designer what you need the design to accomplish, who it should appeal to, what market segments you're going after, who you're competing with, etc. Do not tell the designer what it should look like. You hired them to tell you that. That's what they're presumably good at. Let them do it.

Designers have varying levels of skill at coaching clients through this process. They also, of course, have varying degrees of skill at transforming your needs into a design (as opposed to just pushing pixels into whatever their personal taste is.) If it's not working out, try a different designer. But don't try to be the designer. You say there are six people making "design decisions" -- that should be six people with input on the selection of the designer and evaluation of their work. (Well, ideally it should be a lot fewer than six people, but you have to work with what you have.)

When evaluating the work, try not to evaluate it based on personal taste, but on whether it meets the criteria you set out for the designer to accomplish. (It's a fuzzy line, of course -- there's no way to absolutely eliminate aesthetic taste from what are ultimately aesthetic decisions -- but approaching the task with this intent can at least reduce the number of arguments over "I like pink! No, I like blue!" kind of stuff.)

Democratic-sounding processes such as Etrigan describes above are the second-worst way to get a good design -- there's a reason "design by committee" is an epithet. (The worst worst way to do it is to let the CEO make the decisions. I don't know what it is about CEOs, but every time I've seen one of them take a personal interest in a design decision the design goes straight to hell.)
posted by ook at 9:48 AM on May 23, 2012

A last tip -- the farther I go into my career (yes, I'm a designer, surprise) the more I'm a fan of incremental redesigns rather than radical redesigns.

If you plan on changing everything all at once, well, to some extent you're rolling the dice. You may end up with something wonderful, but you have an equally good chance of inadvertently removing aspects of your current design that work well, or are there for a non-obvious reason. With an incremental redesign, you can focus on elements that are clearly not working, correct them, see how things stand, and iterate. There's less risk overall; the decisions are by definition more bounded and focused; there's less chance of losing brand equity (if we're talking about logo design) or confusing existing users (if we're talking about website design); and at each stage you limit your exposure to the aesthetic whims of either a bad designer or a bad committee member.

(This comes up in software design as well: it's generally considered better to gradually refactor old code rather than to rewrite it from scratch... all those crufty-looking old patches were probably put there for a reason, and if you do a wholesale rewrite you're likely to drop a lot of edge cases on the floor.)
posted by ook at 10:02 AM on May 23, 2012 [1 favorite]

I am a graphic designer as well as a design educator and I agree with what ook, violetk and others have said regarding the fact that you need to allow the designer to be effective at her (or his) job.

They have already outlined some hows and whys. People have also suggested that design is a combination of semi-objective aspects like content, market, positioning, etc. and semi-subjective aspects like color, form, typography, layout, photography, illustration, etc.

Let me suggest that you and others in your organization think about why you have a flat hierarchy... Does everyone have to agree with every decision that is made? If your office went out to eat, would everyone have to eat the same food? Should everyone go see the same movie as an office bonding exercise? Or, perhaps you have a flat hierarchy so that everyone can have opportunities to contribute and be considered valued in the organization. I think there is a difference between valuing people and fostering opportunity and making every decision a consensus.

I am trying to come up with some thought experiments just to push this notion that design should be made by a committee. A good designer will be able to understand the group's vision and provide visual materials that help the group accomplish goals (perhaps even goals the group didn't realize it could have, if things go really well). But, there is a critical element of vision in the design process and vision doesn't come from committees.
posted by Slothrop at 10:03 AM on May 23, 2012

Here is a way I ran a meeting in a hierarchy-less organization where we had to make a difficult decision - this might be of use if you're informal/nonprofity:

After whatever intro was needed...

1. Structured talk with a partner about what you hope and what you're worried about with the issue. This lets everyone organize their thoughts and blow off steam. It also lets the "I must talk" people talk a little. I ran this as one minute per person to talk and then one minute per person to respond. Structure is important so that everyone gets a chance to talk.

2. "Pop-corn style" (people talk as ideas occur to them, not as a go-round) info dump where everyone shares new things only - if they have not shared their opinion before, it's okay to do so now, but remind everyone that this is really for new information and opinions. This is NOT for discussion; it's only for sharing.

3. Review or brainstorm "what is this logo for". Again, ONLY ideas, no responses or debate. List all the things the logo needs to do, and for whom. (We didn't do this as our issue was different, but it's how I'd handle things)

(If you have pre-existing proposals - Use your pre-existing design proposals (which you need to have before this meeting). For each one (and I'd suggest winnowing it to three or four by vote prior to the meeting) do a "pluses" and "minuses". No debate, just a list. (Design proposals can be actual sketches or descriptions or directions - modern, black and white, etc.)

4. Clarification questions only.

5. Either vote on the logo proposals OR make a list of needs for the logo based on#3.

See, here's the thing. IME, most people feel pressured to have strong opinions in these situations. The whole process of choosing generates strong opinions about things people might not ordinarily care about. Then they pick sides, or worse, act out their pre-existing relationships via the thing in question. So there is argument over small differences and trivialities.

The process I am using more and more is designed to get people to talk without arguing. That sounds extremely new age but I have found that consensus often emerges based on facts, or at least you can readily get down to the two best options. Those things emerge from the information-sharing - people take in ideas that had not occurred to them and see problems that they had not noticed.

So this meeting structure is two-fold: let people talk so that everyone feels heard (the emotional/relational side) and present as much information as possible.

It takes a while, but it cuts down on conflict a LOT.

Also, remember to set times for each section - you don't need to stick to them, but you can lead in by saying "and now we're going to spend twenty minutes [talking about THING]" and then at 15 minutes you can say "we have five more minutes". Just as when you alert a toddler to impending changes, you'll find you get better results when you scaffold what's going to happen.

Honestly, running meetings is a skill. Many clever people believe that they can let meeting happen naturally or that all the participants will automatically participate in the most useful way possible - this is so not true!.
posted by Frowner at 10:10 AM on May 23, 2012 [5 favorites]

(To clarify - in the first part of the meeting, people work in pairs, talk uninterruptedly for one minute, listen for one and then have one minute to respond.)
posted by Frowner at 10:12 AM on May 23, 2012

People should not be communicating preferences to the designer. They should be communicating goals. Then the designer can do his/her job, and your team's responsibility is only to make sure their solution accomplishes the goals that have been outlined.
posted by anildash at 3:22 PM on May 23, 2012

Oh boy.

"Design by committee" is a pejorative for a reason. Design decisions should really be made by consumate professionals with input from others, not the other way around. Design is, IMHO, one of those fields where brainstorming doesn't work so well. If you have a talented designer on your team, let them own this project. It might be a hard sell in your flat organization, but if you've got someone with talent for design, you should really let them run with it.
posted by deathpanels at 7:19 PM on May 23, 2012 [1 favorite]

The best practice is to involve everyone ahead of time develop a brief and agree on a set of design principles, i.e., a successful logo design would achieve "x" or have "y" qualities. These are the goals anildash mentioned. This should happen before you consider how anything will look. For example, you should know whether you are projecting a "grassroots" or a "corporate" sensibility, you should have a clear sense of who your target audiences are, and you should explore some metaphors that you think represent who you are as an organisation. Many organisations are very unclear about their own identity, so it's a good exercise for everyone to participate in, revisiting your mission, etc. A good brand designer will do some inclusive exercises to explore the meaning of the brand with all the stakeholders. You also need to think about where and how your logo get used. Do you need something that scales up really big, or has the ability to be recognised in a very small format? Once you've all agreed the parameters, then you can let the designer do their job. All of the primary inputs should come early in the process. By the time the designer delivers their solution(s), the conversation should then focus on "how well does this design achieve the principles we set forth?" That way the conversation is not wide open for debate, but controlled against a pre-agreed set of criteria. It makes decision-making easier, and everyone feels invested in the outcome.
posted by amusebuche at 5:25 AM on May 24, 2012