Great Numbers, Not So Great Design

Let me admit a real prejudice that I have, and maybe you can try to convince me that I’m wrong: it’s my belief that you just can’t get great design out of a design agency with a staff larger than a dozen or two. Design doesn’t scale well, in my opinion, or at least it doesn’t do so easily.

This craft, and whatever pretensions to art it can pull off, rests so much on the efficiency of transferring ideas from the brain to the hand. This means that in its ideal form, it works best when practiced by a single person. The perfect design staff is a single designer who can conceive of and execute an idea from start to finish — a straight shot from the right brain to the wrist — maintaining the same coherent creative vision throughout.

Of course, as an economic matter, this is impractical. For design to work as a business, it almost always has to scale to some degree. The smaller the scale, though, the more efficient the practice of design; transmitting ideas among a small number of people is much more effective than transmitting them among a large number.

Bigger Isn’t Better

When a design operation scales up the equation becomes much more diffuse. Beyond a certain point, a business of designers is no longer a studio — focusing on a specific niche of design, or devoting energies into a small number of projects at once — but rather an agency — a provider of multiple services, staffed by different kinds of specialists. Ideas must travel more complicated routes from brains to hands, and reconciling conflicting signals becomes difficult.

It’s certainly not the case that agencies are inherently staffed by inferior designers. That’s not what I’m saying, let me be clear. In fact, I’ll freely grant that designers employed at agencies are very often more talented than those employed at in-house design groups (except for those in my group, of course).

The problem is that the structures of most larger design businesses cannot effectively facilitate the the transmittal of ideas. They don’t allow good design to happen, because they are overburdened with the organizational overhead of running a business: org charts, jurisdictions, inconsistency, poor communications, etc. All the complications that large groups of humans create for one another when they work together, complications that are not about doing design.

Bench Warmers

When a client hires a design vendor, they are often attracted to a concept called “depth of bench.” This means that the vendor, whether a studio or an agency, has enough staff at the ready in the case that one of the staff assigned to their project fails miserably or just leaves the job. To provide an attractive depth of bench, a design vendor almost invariably has to scale up.

What you get, then, is a grab bag effect. Hiring a design agency with a large staff is an exercise in luck of the draw; one client may get very good designers assigned to their project, and another may get poor designers. Most often a client will get a mix. A random mix. In a large design agency, I contend that it’s impossible to have a staff that’s uniformly excellent. Impossible.


By contrast, design studios that are small in staff are also much more focused. Their structures are inherently simpler, so their dissemination of ideas is more effective. Moreover, in many design studios where it’s clear who provides the source of creative direction, there’s very little ambiguity regarding who is the brain and who are the hands. (Contrast this with agencies where there are often dozens and dozens of dubiously talented people with the title “creative director” printed on their business cards).

These small studios are usually good at a fairly narrow band of project types, too, though they may have delusions of being adept at a much wider spectrum. You wouldn’t hire a Web design studio to design your outdoor advertising campaign, for instance, nor should you. As a client, you want the very best design you can get given the medium in which you’re dealing. Do you want to hire a company that does a few things very well, or a company that does many things not particularly well?

  1. I completely agree. So much of what we do as designers comes from instinct and even when you have a group of uniformly fantastic designers, getting those ideas across to your collegues is often a futile exercise. There is no better way of designing for me than being allowed to let the tasks at hand become my single and sole occupation, and trying to explain decisions, or delegate tasks does nothing but get in the way of producing what I have in mind when that eureka moment strikes and the whole thing becomes clear. GANNT charts and flow diagrams are great when your boss is over your shoulder asking for a progress report, but it just serves to dilute that clarity of thought you have at some point in every successful project.

    I guess what makes a great team is one which is able to communicate effectively each members own eureka moments, and one which allows everyone to work their mojo. Which is why I’d agree that the old adage ‘too many cooks spoil the broth’ is as apt in a web design house as it is in a kitchen.

  2. Even in a small firm like mine (6 people), you can run into issues… the money person controls the communication which handicaps the creative person. The messages get mixed or poorly communicated. And I *like* who I work for.

  3. I’ve worked in agencies both small, medium and large, and on the client side, and I agree. Larger agencies, in my experience don’t produce consistently high quality work across all of their projects, and don’t always offer their clients value for money. They’re also not necessarily pleasant places to work.

    The type of organisational structure that works okay in a small agency very quickly becomes unwieldy when you get up to sizes of more that 20-30 people. An agency with 60 people typically won’t have many more ‘doers’ (creatives, designers, developers) than an agency half of that size.

    Instead they inject more process into their projects. Big agencies inevitably charge more as they seek to turn a profit in excess of their large wage bill. Clients are paying more for projects so they need to see more demonstrable evidence of expertise. This turns into workshops, focus groups, user personas, IA studies, wireframes – a process that can last a number of months before any actual design begins.

    There’s nothing inherently wrong with any of the above practices, the problem lies in the fact that most agencies operate on a short-term, project-based workflow. Their focus is on getting the current project out the door, and billing the client for it. Once the project has passed through the studio, it’s forgotten about and the agency moves on to the next priority. This means that learnings and insight that result from the research and development at the beginning of a project rarely have their effectiveness measured over the long term.

  4. Design was never meant to be made by committee. I remember working in previous agencies and the greater the number of people in the meeting, the worse the idea came out in the end.

  5. I’m glad to see that other people have the same ideas! I’ve staked my business on building a small, focused design studio rather than a large agency. Clients certainly seem to agree, and there’s a groundswell here in Portland, OR, around building small agency partnerships to provide the best solutions to clients.

    Keep it small, keep it smart, keep it collaborative.

  6. I agree with your critique of the burdens of large organizations, and I think the same thing could be said for many creative tasks. However, certainly there is a collaborative aspect to creative projects that can be very rewarding and, in the end, lead to excellent results. How does that fit in?

  7. Nice encapsulation of why Polychrome is small. We play as a band, and we key of each other, without the need for everyone to solo or for one person to carry the lead on every song. Very, very, very hard to do once you get past a certain size.

  8. I guess I’ll be the lone voice of dissent here. I dont really think it matters all that much. I’ve worked on teams from 4 to 70. I’ve seen good and bad work come out of both. In the best cases, I’ve been able to spend more time creating, less time dealing with things that dont relate to making things, and been able leverage things like actual user research that I never had time to do in a tiny group.

    I’ve found the most important thing is the frame of mind of those people tasked with actually running the shop.

    So any way – this is all just a long way of saying I think there are way way more important paths to success than size of the organization. Or rather – while we have a case of correlation, I see no evidence of causality.

  9. This is an interesting idea, and I enjoy working in smaller groups, but in terms of the success of the end product, I think the attitude of the client is more important than the size of the studio. If the client doesn’t understand the process or wants to drive the design, it doesn’t matter if there’s one designer or 100… the results will suffer (and so will all involved, including the client himself.)

  10. I don’t know. Sagmeister was making the same point a while ago and it seems to justify his own existence but… I’ve seen some good work from Huge and Hampf, I’m sure there are others.

  11. I’m going to dissent, but just a little.

    First of all, I agree that ‘scale’ is a omnipresent problem for organizations, and have informally observed that groups don’t function particularly well beyond about 30-50 people.

    But I think that a major aspect of the original point was more about the size of a team working on one project. Here, I agree that somewhere between 2 and 4 is as big as you should go. I don’t say 1, because I think that discussion and multiple viewpoints can always benefit the output — especially when the team is a multidisciplinary one.

    MAYA Design (disclosure: I work there) provides an excellent model for how to work within both of these constraints while still enjoying the benefits of a larger organization (multidisciplinary, collaborative teams; different skill sets; load-balancing). First, we’re only about 40-some odd people. That means that we can function relatively efficiently with ‘minimal management overhead’. Second, everyone is a practitioner. This means that we are all responsible for finding projects and working on them. Paired with a ‘you kill it, you eat it’ habit, clients almost never experience the pot luck syndrome — the people you first meet with will almost always be the same ones who tend to your project.

  12. There’s that old adage, ‘keep it simple stupid’.

    The more people you introduce – the less chance you have of high quality work.

    Dave does work that ranks within 70%

    Rick does work that ranks within 80%

    John does work that ranks within 100%

    Total Team Design Potential: 83%

    Total Team Project Time: 33%

    John Design Potential: 100%

    John Project Time: 100%

    Quality – vs – Time

  13. Designers employed at agencies are very often more talented than those employed at in-house design groups.

    Uhmm, I dissent about this line. As I’ve myself been working as an in-house designer. In-house designers’ works are limited and confined in the certain field. I don’t know what things made you write this line. Certainly in my case, I don’t have variety of projects every 2 or 3 months, I have to keep working on our own products and internally stuff. of course some of my work never get out, stay indoors.

    Apart form this; I think most of problems here are management problems. If you are dealing with the best project manager or creative manager, projects wouldn’t loose quality. Numbers don’t matter if you have a good lead. There may be frailer of quality management in the organization.

  14. I always give clients who are leery of working with a small studio (or a lone designer) a metaphor for the small shop/single designer experience:

    A designer is a chef. The client is the diner.

    Diner tells chef: make me a four-course meal.

    Chef and diner then discuss what they’d like that meal to be, what the diner’s tastes are, how it meshes with the chef’s style and competence.

    Chef goes to buy ingredients (sometimes the diner comes along, or has already brought the ingredients with them. Interesting restaurant, eh?)

    Chef retreats to kitchen. Cooks.

    Presents meal.

    Diner eats.

    If the chef and the diner have chosen each other well, then the diner should leave satisfied.

    Perhaps a strained metaphor, but for me, the content are the ingredients, and the designer is the chef who puts it all together to make something palatable.

    And we all know the old saw about too many cooks. If you want a (perhaps) more predictable meal, go to the Olive Garden. Or Burger King. You’ll get served faster, but your meal will taste a lot like the one served at the next table. And it won’t be made just for you.

  15. Very intersting metaphor Alan. The ‘Burger King’ agenices tend to do exactly that – do a little bit of the burger then pass it on.

    Want extra cheese? That will be another $500 thanks, and we’ll pass it onto the cheese guy. Whereas the Chef designer won’t mind throwing in extra cheese as long as it doesn’t take much time to source.

    Ok getting too carried away with this one. 🙂

  16. This means that in its ideal form, it works best when practised by a single person.

    I find that a very odd thing to say. Maybe for logo design or something very specific, but I know that I’ve personally found all the most successful projects I’ve worked on to be the one’s where I’ve collaborated with other excellent creative folk.

    Sounds to me like the issue isn’t fundamentally down to numbers, but to leadership, and that the greater the numbers, the harder it is to have strong design leadership.

  17. I used the metaphor of a band deliberately above, and I think it’s an apt comparison to larger organizations as well as to solo artists. There is a *definite* advantage to being in a right-sized ensemble. You feed off each other; when one player is down the others can carry the load; when one player raises the bar the others get more out of themselves. Just as in music it doesn’t mean there aren’t brilliant solo performers, but it’s in an ensemble setting where you usually see the real magic happen — and that doesn’t mean that everyone has to contribute equally. My rule has always been when the band gets so big that someone needs to stop playing and be the conductor you’re not a band anymore.

  18. I feel compelled to respond, but I don’t know exactly how, because it’s unclear to me what the point of the original post is. It’s also not clear to me what Khoi means by ‘design,’ because this statement makes no sense to me, in my work:


    This craft, and whatever pretensions to art it can pull off, rests so much on the efficiency of transferring ideas from the brain to the hand. This means that in its ideal form, it works best when practiced by a single person. The perfect design staff is a single designer who can conceive of and execute an idea from start to finish a straight shot from the right brain to the wrist maintaining the same coherent creative vision throughout.


    It has been a long time (and possibly never), since I worked on a design problem so simple that it was ideal for a single person to handle it. A key reason for Adaptive Path’s scale has been the complexity of the design challenges we face. In order to tackle complexity, you want people with different skills, backgrounds, and perhaps most importantly, perspectives.

    And I know you know that Khoi. I can’t believe you think would be better if you could somehow handle it all yourself.

  19. I’m all for a small design studio who have fantastic designers in their respective discipline working together. Everyone wants to be the best at everything, but reality is that, not all can.

    For me, I rather be the best at what I can do and collaborate with other fellow friends/designers who I feel are fantastic and fits the bill for a project.

    In that sense, design studios should employ such similar method. To me, collaboration is the key in design world. Unless you’re a rare super talented master of all trades of course.

  20. Apart form this; I think most of problems here are management problems. If you are dealing with the best project manager or creative manager, projects wouldn’t loose quality. Numbers don’t matter if you have a good lead.

    I have to agree with Carla Pinle here.

  21. Designs scale well if an entire staff worked in a higher resolution. That’s probably why I wouldn’t hire a Web design studio to design an outdoor advertising campaign. Because they’re always working in 72 dpi. 🙂

  22. Peter, yes, I know that most design can’t be done by one person. I’m certainly not advocating that a 1-person shop should be the ideal that design businesses aspire to. That was just the basis for my premise that design gets harder as more people get involved.

  23. Very provocative premise.

    The problem with small design companies is that the whole group could be pinheads. At least with a larger agency there’s a good chance a few of the people are going to be worth their salt.

    I make websites in my solo shop and I think the results are pretty good for my clients. With that said I’ve streamlined what I do down to specializing in RadiantCMS site for small businesses because I’ve got to seriously focus in on niche so I can really master my product.

    Meeting clients, discovery, architecture, compositions, css and xhtml templates, custom programming, configuring servers, and developing an effective marketing strategy is a lot for one person to do. I know I could be better at design if that’s all I did. I could be a better coder if that’s what I did. I only spend like 25% of my time in any one of the disciplines needed to produce websites.

    Advantages of being a solo act are: when I’m meeting with my client I truly know my limits; in all phases of development I’m able to angle everything towards my skills; when making a comp I’m able to plan the coding and programming needed to make it come alive: when coding and programming I have the freedom to make changes to the design as I better understand the ideas in the comp and my client’s needs.

    Khoi, I appreciate your point and I’m glad you said it. Yeah, the perfect mix of super talented designers, programmers, writers, marketers, and management could make mind blowing stuff. But that doesn’t happen all that often. What Khoi said is what mostly happens. You’re dead on.

  24. I think there are two key words to follow when concerned with design groups. That’s business and design. For businesses to truly flourish, they flourish because they have good management. Successful companies with groups, big or small flourish because they have great managers. Even successful one man shows flourish because they too have successful managers.

    For design to work as a business, the design has to work. Plain and simple. There are so many issues in design to deal with. You’d think that in this age, technology would have closed the gap between ineffective communication.

  25. I couldn’t help but notice one of the open position calls for a Managher… I am sure no one would make a typo at the NYT, so I guess a managher is a new and improved position above the manager level?

    Just kidding 😉

  26. @Tom Dolan – ‘My rule has always been when the band gets so big that someone needs to stop playing and be the conductor you’re not a band anymore.

    Indeed – that’s when it starts being an orchestra!

  27. Too many Chefs spoil the broth.

    I haven’t read all the remarks above, and someone might have already mentioned this. So I apologize if I am repeating someone’s remark.

    Here is another one for ya.

    ‘A board never reached any decision.

    Even a decision that they reach, is still an average, and a compromise. Hence, it is not a decision, atleast in my book.

  28. > that’s when it starts being an orchestra!

    Agreed. So what’s important to note there is orchestras perform differently than bands. Orchestras don’t jam. Orchestras don’t improvise. Orchestras don’t go off of the script, but rather perform according to direction. Individual players are encouraged to excel and express themselves, but within very proscribed parameters. Orchestras are able to deliver complexity, precision, and are incredible ‘systems’ at processing a well-designed plan. If the plan is so scripted, they can frame and deliver wonderful moments of individual virtuosity. The question is how closely does this resemble addressing the design problem one wants to solve?

  29. If talking music, They are like the five music notes; combinations that produce endless melodies, like the five colours: mixture that produces a variety of beautiful objects.

  30. My experience is that design quality is highly correlated with a design team’s collaborative health more often than it is with the size of the firm in which the team works. Big firms can have highly-functioning teams; solo practitioners can be inept; small firms can be dysfunctional. (Disclosure: I work for a small, functional firm — Cooper).

    As many here have pointed out, collaborative health is not something that just happens when you throw designers together. A highly-functioning design team is the result of conscious effort, and its creation is a design problem in itself.

    A healthy team subtly blends a variety of organizational imperatives — effective project management, strong design leadership, deep and clarifying problem-definition, sparky and elegant consulting, and so on — along with great designers with clearly defined roles. Using the Utah Jazz of the late 90’s as an analogy, someone’s gotta be John Stockton, someone’s gotta be Karl Malone, someone’s gotta be Jerry Sloan, and none of it works unless everyone holds down their responsibilities.

    FWIW, I can’t imagine many projects (that I’d want to work on, anyway) in which the ideal is a ‘single designer who can conceive of and execute an idea from start to finish.’ To solve a problem of any complexity, I’d rather have effective thought partners who share my approach than attempt to instinctually avoid rat holes, extraneous details, and the juicy (but irrelevant) periphery.

  31. Working with a ton of designers … Well, sometimes all you can do is shrug and say ‘At least s/he makes me look good.’ Just getting to know people and share your opinions on design is more difficult and more scary as a person in a bigger people. It’s very hard for people to function in a large group.

    There’s so much to be said for having a singular vision in design to unify behind. It’s hard to get a large number of people to have a singular vision for the organization’s design mission

  32. Being on a team requires humbleness and the ability to compromise even on ideas you may be married to. In my experience, many if not most designers completely fail at this aspect.

    Software companies are something you should look to emulating when getting to a large size. Often projects and contracts are given to an internally formed group of programmers and software designers no larger than 5 to 6 people.

    Of course designers for the most part refuse to accept that software design is design at all.

    I submit this comment on behalf of all software designers, they that get none of the credit or recognition when they create a masterpiece but receive all of the blame when anything is wrong.

  33. I am responding to Khoi Vinh comments about the size of a firm. I wonder how he feels about the relationship between the size of the problem and size firm.

    Quality of design is an illusive concept. One might suggest that quality design is the ability to deliver effective and beautiful solutions to wicked problems consistently.

    His comments have me thinking. Having scaled back the design group at our firm to 15, I often find myself longing for the ‘deeper bench’. I miss the ability to prototype quickly and having a wider range of talented people contribute in surprising ways close at hand.

    At the same time, I do think that the most beautiful work in history has been typically made by individuals.

  34. My final comment is a job post in an unlikely spot — but if you’re reading Subtraction you’re likely my kind of designer. We’re looking for a couple great players to add to our ensemble, so we can be that perfect size. : )

  35. Once upon a time, not so long ago, your personal blog seemed to be just that. A large percentage of your recent blogs seem to be NYTimes-related. Well, that new building is lovely. It must be like walking into a gigantic radiator, like a post-modern version of Eraserhead … *LOL*

    Many commenters of this post have presented a variety of metaphors (some more applicable and entertaining than others), but I don’t think anyone wants to be expendable. My guess is that business at The Times is driven by the paper’s content, which attracts the ads that pay for, well, more content. Yes, also nicer buildings and other businesses, but primarily more content. The writers and photographers (some more popular than others) create the content that appears in the paper and on the site. There will always be news, so there will always be a need for people who can relay the news. What happens once the perfect facsimile of the paper is created on its website? That corps of talented but faceless and nameless designers is no longer needed.

    Khoi, this is the 7th version of your blog. is all you, from the rooter to the tooter. Here you are free to be as bleeding edge as you wanna be as both its primary content creator and its sole designer. The current design of works because it replicates so much of the paper’s original design. Yes, there are ancillary projects (like the redesign of that job board), but once that optimal design has been achieved – then what? (Okay – obviously – but then what?) It’s easy to scale the role of designers within that context because design is more than likely not what keeps the lights on at The Times.

    Still, there must be a way to scale other aspects of attracting and maintaining a great design team there – especially since The Times is no longer just its paper, but also a conglomerate of other media properties that bring in revenue. (Which kinda makes its design team look more and more like an agency, but anyhoo …)

    Say, for instance, that The Times bought Yahoo! now that it has become more affordable. Whether or not that purchase is a great idea for your company is beyond your control, but it happens and the redesign of it lands in your lap. How would you measure the interest of your team in taking on a project like that? You’ve developed a smaller, more effective team with a certain focus, but what if no one on your team was particularly interested in that project – a project that would probably be substantially different than what attracted them to The Times in the first place? How would you relate the scale of maintaining a small in-house design studio to the scale of the growth of its enormous parent company – or would you?

  36. It’s my belief that you just can’t get great design out of a design agency with a staff larger than a dozen or two.

    Actually a lot of people in the video game development industry feel the same way. I remember watching the IGDA (International Game Developers Association) conferences videos a few years back and this kept coming up. Development teams of 25 or less were optimal, efficient and highly connected amongst each other. In effect, there was more passion, connection, and ownership with what they were doing. As soon as you got above the 25 people mark, a disconnect started happening and there was less passion (i.e. ‘just doing my job’).

    ‘The problem is that the structures of most larger design businesses cannot effectively facilitate the the transmittal of ideas.

    True but it’s not so much the size as the process. Businesses today just follow the cookie cutter approach to how they should run their business, especially as it gets bigger, and this more often than not destroys the great startup culture that exists when companies are small. Businesses don’t have to fall into this trap. There are ways to maintain this startup culture as you get larger but it requires a different way of work / process (a more natural one). Two examples are below.

    The Connection Culture by Michael Lee Stallard

    Small Giants by Bo Burlingham

  37. I would argue that the studio of Art Lebedev produces excellent works and is considered a ‘Great Numbers’ studio.

  38. You should all spend more time creating design and less time typing unoriginal, long-winded philosophies. Yawn.

  39. I my experience both as a sole developer and working with a large agency I think the difference is not necessarily size – it is about project management.

    The challenge is that all good designers are fiercely

    proud of their skill and are determinedly individual.

    *Managers* however always speak the language of inclusiveness, fudge and compromise. And by definition if we agree that all the sub sets of skills involved in web development are specialisations, then it follows that the *manager* will know a lot less about many of those disciplines

    than the people involved in the project. Generally though the big barriers to developing an effective net communications strategy are the clients and our in house account managers 🙂

Thank you! Your remarks have been sent to Khoi.