fbpx Techne | HowlRound Theatre Commons


Goodbye Waterfall, Hello Agile

We often consider technology an imposition on both our craft and our business: social media putting pressure on everything from marketing to performance, for example, or IT budgets draining resources from “more important” work. But the etymological root of technology is techne: the Greek word for art. To paraphrase cartoonist Walt Kelly, we have met the enemy, and it is us. Every month, this column will investigate the ways in which technology can inspire us, transform us, and help us chart a new course in the 21st century. Thanks for—to use a radio metaphor—tuning in.



How often do you stop to think about the metaphors that govern the way we make theater and run theater institutions? The “hidden patterns” we follow, often subconsciously, as we structure our work? There are two such hidden patterns that I think we really need to pay attention to. I’m going to call them waterfall and agile, for reasons that may be obvious to anyone with software development experience… because that’s the milieu they come from.

In waterfall software development, a project moves through a sequential series of phases: roughly speaking, definition, design, development, testing, and maintenance. The basic idea is that the thing being built, whatever it is, “pours over a waterfall” from phase to phase. When the product has been defined, design can begin; when the designs are approved, programmers begin development: you get the drift. The whole effort moves in one irreversible direction. You never turn back.

Waterfall is a clear tried-and-true methodology, with its origins in the American manufacturing industry…but it comes with significant limitations. The big one: linearity. If you happen to have an insight about how a piece of software ought to have been designed, for example, when you get to the testing phase, you can’t make changes—not without running through most of the process again. The product you get at the end of a waterfall process, therefore, is almost always exactly what you set out to create at the beginning…but by the time you’re done making it, you often want something else instead.

This is, I think, how we make a great deal of our theater. A playwright writes a play, then she hands it off to other artists to design and build; the plays gets tested in front of preview audiences, then launched to the world on opening night, then maintained until the run is over. I’m (obviously) simplifying, but the waterfall parallel is still (I think) sound.

This is also, in many ways, how we run our theatrical institutions. We conceive of an entire season, then we implement that season, then we launch it and see how it does. We develop waterfall marketing campaigns, we sell waterfall tickets (bought far in advance of the intended date…by which time a theatergoer might not feel like attending any more) and we staff our organizations to support the waterfall methodology. We are governed, in the main, by the waterfall metaphor.

But waterfall isn’t our only option. We could make agile theater instead.


It’s time to start paying a bit more attention to those entrenched patterns we’re relying on…and deciding more overtly which might be serving our sector the best.


In agile software development, a product is built in a succession of slightly-improved drafts, rather than in carefully aligned phases. You make a rough version of what you want to build, then you show it to some of the people who might use it and gather feedback, and then you use that feedback to make the next version. Lather, rinse, repeat. There isn’t one big ejaculatory launch at the end of development; there’s a series of smaller rolling launches, each slightly more robust than the one before it.

Perhaps more importantly, in agile development a product is built by a self-organizing, cross-disciplinary team working in parallel, rather than by a series of expert solo practitioners each doing their jobs largely by themselves and then handing their work off to the next expert in line. Every team member contributes his or her own expertise to the general effort, but nobody’s perspective is sacred; everybody works in service to a product’s users. To rely on a sports metaphor, if waterfall development is like a relay race, then agile is more like a rugby scrum.

The primary benefit of agile development? Adaptability. If something isn’t working, you find out right away and adjust before it’s too late. The difficult notion for some people to come to grips with? That you don’t have a product entirely planned out, down to the last detail, before you start working. When you set out to make a piece of software, you never know precisely what you’re going to end up with—a feature that might seem obvious or essential at the outset of a project might not “test” particularly well with your intended users—but do you know people are much more likely to love it.

Increasingly, we are making agile plays, too. The obvious example: the work of devising ensembles: interdisciplinary groups of theater artists working collaboratively to create plays. (Look for a future column about agile and devised theater.) But I think it’s also worth asking what an agile theatrical organization might look like, too. A few thoughts:

  • An agile theater wouldn’t have seasons; they’d just develop plays and release them when they were done.
  • Speaking of which: an agile theater would ONLY do new work. (Waterfall is much more suited to well-traveled plays.)
  • At an agile theater, rehearsals would be open to audience members on a semi-regular basis. (How else to get their feedback?)
  • At an agile theater, audiences would be deeply invested in the work that ends up on stage, having played a real role in its creation.
  • Agile theaters wouldn’t sell subscriptions. A subscription is just a kind of “theater attendance blueprint” to be slavishly followed. They might, however, develop loyalty programs to reward recurring attendance… and tickets would always be exchangeable.
  • Agile theaters wouldn’t necessarily be run by artistic directors; they’d be more likely to function as collectives.
  • If an artistic director did run an agile theater, her job would be to select devised projects, rather than plays.

The implications are wide-ranging and significant.

And all of this is happening, right now, throughout the American theater. With the rise in devising ensembles, the emphasis on audience integration, and the decline of season subscriptions in some quarters, we are arguably at the beginning of an agile theater movement. A significant percentage of the small independent theaters in my own city, DC, would definitely fit the agile description—Rorschach, for one, has abandoned the notion of a “season” entirely; dog & pony, for another, has been bringing audiences into its devising process earlier and earlier—and I suspect the same is true elsewhere as well. So I think it’s time to start paying a bit more attention to those entrenched patterns we’re relying on…and deciding more overtly which might be serving our sector the best.

Bookmark this page

Log in to add a bookmark

Gwydion Suilebhan discusses the impact of technology on theatre.



Add Comment

The article is just the start of the conversation—we want to know what you think about this subject, too! HowlRound is a space for knowledge-sharing, and we welcome spirited, thoughtful, and on-topic dialogue. Find our full comments policy here

Newest First

As a playwright AND software engineer, this post was great to read. Having worked on both waterfall and agile projects in software development, the analogy really rang true for me.

Seeing Neal's comment below, I was reminded how people have often been hostile to new ways of doing things. Nonetheless, these comments are not without merit. When it comes down to it, you can make crappy work with both waterfall and agile methods. And also, no single methodology is right for every single project. You need to make a decision based on the project, the collaborators, what you want to produce and, most importantly, for whom.

I agree with Catherine below. Agile methods actually require MORE rigor and discipline than traditional methods. When people are working together and there's no formal blueprint at the beginning, you need to work harder to make sure you're making something of good quality.

How to engage and audience and get feedback from them (amateur or not) if you're creating work in smaller iterations? This is a really interesting question and one that seems tricky to answer. There's only one way to find out, try different things and see which one works.

What really rang true for me here is that sometimes during performance, I make discoveries about my work (usually bad) that would only have come out in front of an audience and wish I had the opportunity to go back and work at it. This way of thinking really addressed that nagging feeling.

Thanks for starting this conversation!

This is a really interesting way to frame different styles of creation and administration in the theater. In my own experience, I see companies struggling to overlap waterfalling and agile models--sometimes successfully, sometimes not. What is "development hell" but an unsuccessful attempt to insert an agile process into a waterfall organization?

I must cop to my own snobbism here: as a playwright who has invested significant time and energy into my craft, I am not always delighted by amateur suggestions. There. Said it. I AM open to hearing what's behind the suggestion: this ending is unsatisfying, or this character confuses me, or I don't like any of this. Or, simply, "I went to college and I need to use the phrase dialectical materialism out loud in this room."
Waterfall can get stuck, get stale, get ahead of itself, and fail. Agile can fritter away energy, empower the loud above the thoughtful, and can make the office politics of a waterfall organization look like reindeer games. Garbage in, garbage out, to continue the software metaphor. The best devised/agile work I've seen (and I love it) is done by companies that are really rigorous, really thorough. It's impressive. It's not some wishy-washy stone soup. It can be as high quality as anything out there. Just as waterfall-model work can, and a waterfall organization can succeed brilliantly. It's about finding the shoe that fits, always.

Though I agree with parts of this essay (such as in terms of getting rid of subscriptions in favor of better methods of encouraging attendance), I respectfully but fundamentally disagree with the basic thesis (in regard to play development), starting with the very metaphor, which I think is false.

I suggest a better metaphor would be a general contractor trying to construct an original building without a valid blueprint, insisting that all of the craftsmen are so skilled and experienced that they can just wing it.

Would you buy a house from such a company? Or would you want one that starts with proper blueprints, because the craftsmen have enough pride in their work that they don't feel the need to do the architect's job as well? (Just as the architect respects the craftsmen [and women] enough that he [or she] doesn't try to do their job for them.)

"If it ain't on the page, it ain't on the stage." Many theater people seem to have forgotten (or simply don't believe in) this old saying, but I think it is precisely this "agile" attitude (which as the author indicates has in reality been around for some time and, alas, seems to be gaining strength) that is partially responsible for why so many former theater-goers have (as one previous commenter put it) exercised "quality control" by no longer attending theater.

Too many theater people have the attitude that another commenter expressed, that what the audience sees opening night (and for the rest of the run) doesn't matter, that what matters is "the process". I think this attitude is fundamental wrong (and disrespectful of the audience's time and money) -- and needs to be corrected in a hurry if theater is going to survive.

Perhaps I should have commented more subjectively; that is, just because I seem to prefer plays that were good to go on the page to begin with (and then made great by the actors, director, set designer and everyone else involved with the production) doesn't mean that others aren't free to prefer more agile plays. Whatever works!

I feel like I have a lot to say about this, but not sure I can really articulate it all. First, I'd like to thank you and give my support to an agile mode of producing. I think it contains a number of the ideas I've had in mind, I just didn't have a word for it. Now I will steal yours.

I do want to note that an agile mode can work for classical works, or newer already produced productions. I will say it over and over, but a well-produced well-crafted revival is every bit just as much a new play as a first production. As much as I love playwrights, they are not the only creators and generators of new work. As I see more classical theatre I realize how a talented director, actor or designer can make any play feel relevant and exciting. I think the idea of agile producing should allow any creative to generate a production.

And to Brett's question below, I've always felt the way to invest the audience is to make them familiar with the players and the craft. I agree open rehearsals do not generate new attendees, but seeing a work in development, seeing how they are important to how the artists take their next steps, having conversations and building relationships do grow audiences. And more than that, it deepens them and makes them feel like they are an important part of the theatre system.

One day, Hannah, I'd love for you to expand on the claim you've made here (and that I've seen you make elsewhere) that a revival can be a new play. I just don't buy it, but I'm intrigued by it, and I want to.

I would say that a revival might "feel" like a new play, and might certainly be relevant, as you suggest. And I would agree that an agile-like rehearsal and production process is still not only possible but frequently what actually happens. But if you begin from a set of blueprints -- an existing script -- I think you're always in some sort of waterfall-agile hybrid process. (Which is absolutely fine, really -- no judgment implied by that description.) Unless you're going to significantly revisit/revise that script. Right?

This isn't an argument, more a rumination:
I've seen the notion of "audiences being invested in the work onstage because they were involved in its creation" spoken of a lot. I'm not certain whether this is true? The impression I get is that open rehearsals and the like are poorly attended, mainly by a small core of theatre buffs, and that most playgoers, casual or dedicated, really just want to see a finished work, not the sausage being made.

Really, just asking if this is a notion that's been explored in depth anywhere else/is supported by evidence vs. it being a romantic idea that doesn't really play out. Either way, it doesn't seem like it's in any way *bad* to invite the audience to attend even if they choose not to - I just wonder whether it's talked up more than it bears out.

So I came late to reading this (shame on me) and then found a request for a response. Sorry G that it took me so long. Some thoughts:

- America's cultural participation is moving rapidly toward "make and share" vs "sit back and experience." Yes-- there is a large population who want to experience finished products only, but I would point out this tends toward the older, more traditional theatre-goer and challenge that (regardless of age or prior experience) it's not because they don't want to interact with the creation process.

- Open rehearsals at most theatres place patrons in a similar, if not exactly in the same, position as in performances: a passive one, as watcher or feedbacker, as a body to absorb but not effect. It is only going to be interesting to a select number because the experience isn't all that much different. And what harm is it if it excites people who already like theatre? Now maybe they like your's more.

- There has to be something to invest in. Providing more and more information about yourself may or may not reveal a hook, that "something" for the other person (the audience) to sink into themselves.

- If I only had "audience" that came to all our open rehearsals, testing sessions, workshops, etc and never a ticketed performance--why shouldn't that be as awesome as the "audience" that comes to multiple performances during the run? I love it when people totally new to dog & pony dc come to one of our non-regular-performance events--because their first experience with us has the potential to be more intimate and personal.

Thanks for articulating this Gwydion. It sheds light on a lot of what is already happening in the ensemble field. I think there is a lot to be learned from the agile model of tech companies that you describe above. But I also think we should not throw out the baby with the bathwater in terms of more traditional, or waterfall, producing models. A couple of thoughts to consider, from foolsFURY's experience.

First, I think there's a somewhat false divide between "devised" work and scripted work, or work in which the text is being created primarily by a single playwright. Often we, and other ensembles, use an existing script as the framework around which to build a unique production. This isn't technically devised work, or even "new work" by most standard definitions, but it can be just as time consuming in terms of iterations, solicitation of audience response, and revision. Which is to say don't think one needs to be creating devised work to take advantage of the benefits of the agile model!

Second, there are a couple of pitfalls of the agile model within company organization which need to be addressed within whichever company is undertaking to work this way. For one, there is a need for some specialization in the producing work for an audience, whether in the season model, or less consistent. Not all teams of artists, nor individual theater makers, have the producorial skills to bring the work to fruition for the public (a premiere being very different from an open rehearsal or work-in-progress showing). In the ensemble world we are always looking for people who can be both artists and administrator/producers, but the specific roles are often not interchangeable. It's easy enough, in the for-profit tech world to hire additional experienced project managers to control the reins of specific projects. This is, most often than not, not the case in the non-profit theater world. Some sort of filter (positive term) or bottleneck (negative term) that runs through the producers is usually necessary and expedient, rather than having the participating artists remake the wheel, or climb the steep curve of ;earning to produce.

Perhaps more importantly is the need for what the corporate world would call "quality control," - what I think of as the combination of maintaining the company's specific aesthetic sensibility, and keeping the bar that we are trying to achieve at the same high level. These are both critical, particularly if the company is not producing a regular season, but is sustaining audience interest by presenting a specific aesthetic "brand." Both of these goals can be challenging once artistic and production leadership is shared among too many people. There are great examples of this working beautiful - the Rude Mechs in Austin, Tx, with their multi-headed co-producing-artistic-director model. But there are also examples where, despite the best intentions of superb artists, collective or shared leadership can create uneven quality - for example the final years of the remarkable Jeune Lune, in Minneapolis, in which they shifted back and forth between a joint artistic leadership to sole leadership under Dominique Serrand (this isn't to say that that shifting was responsible for their ultimate demise - but worth noting nonetheless.)

All this said, using some of the strategies of the agile model is extremely helpful, particularly as innovative new works, be they devised, written by one writer (see Taylor Mac's comments), or creative approaches to existing scripts, takes as much time as it does.

Thank you for outlining this for us all so clearly!

Thanks, Ben, for the opportunity to clarify a point about which you are absolutely correct. I can see why you (and others) have interpreted that I meant to directly equate waterfall with plays written by individual authors and agile with ensemble-generated work, but I did not mean to make that assertion. Far from it: in fact, I'm about to begin my own new piece for The Welders (a largely agile organization, I would argue, producing both agile and waterfall plays), and although I will be the only author, I will definitely NOT be creating the piece via a waterfall process.

I agree, furthermore, that there are challenges to running an agile organization -- the greatest of which, I believe, is that having grown up in a waterfall world, we have waterfall embedded in our brains at a deep level, and thinking in a new way is very hard to do. As a sector, we still have a lot of thinking to do about what an agile theater company might look like. To your question of quality control, for example, I would posit that perhaps we need to let our audiences be the arbiters and/or keepers of quality... but to an extent I haven't personally figured out. We clearly have work to do.

And finally: I really do NOT think we need to abandon the waterfall model. It's fine and useful its way. For me, though, it's just so steeped in the language and structure and mindset of capitalism and profit and product and manufacturing that I have a hard time reconciling it with the values I prefer to associate with art. Those may, however, be my own struggles only.

Gwydion - Thanks for more thoughts. I'm especially interested, and in agreement, with your thoughts about the "language and structure and mindset of capitalism...." That's very poignant, and yes, it is very hard to distance ourselves from that capitalist mindset.

I believe that much of what the current ensemble theater movement offers is at least a partial alternative to this with a

greater emphasis on collaborative creation and decision making.

I totally (but respectfully) disagree with the idea that we need to let our audiences be the arbiters and/or keepers of quality." I think we DO need to pay a lot more attention to our audiences in terms of what they are interested in, and what they'd like to see on stage. But we would do them and the field a disservice to make them responsible for the bar of quality. To make a vast generalization, that certainly smacks of a certain elitism (so sue me....), the majority of American audiences are rarely exposed to the highest quality that live theater can achieve. Asking them to tell us what is good enough is either a) likely to lower, stabilize, or slightly raise the bar - but all within the realm of what they know and have already experienced, OR, another perspective: American audiences have already spoken about what they want to bar of quality to be by simply decreasing their theater attendance over the last fifty years. If that's the case, then they are saying "raise the bar if you want us back" - and most of us simply aren't hearing it!

Okay, enough ranting. Clearly I am a proponent of us all making better work, and raising our definition of excellence - my company as much as anyone else. I know this wasn't your main point - but it's a key issue in all of our work.


You know, I completely hear you. And I was, to be clear, wildly speculating with that comment, which was born out of a STRONG suspicion that audience integration is going to mean giving up more control (or the illusion of control) than many of us are going to feel comfortable giving up.

I suppose what I really mean is that, by showing our work in rougher and rougher stages to audiences -- by exposing our process and integrating them into it -- we are going to be paying more attention to what actually engages them than we ever have before. We are still of course going to be the experts -- no one can take that away from us, nor should they -- but what we're going to need to become expert AT is going to have to change (and be less self-centric).

I wonder if that might be a common ground between us :)

Thanks Gwydion, this helps me define what I'm doing with the process of one of my new works. it's not devised but the development is happening over multiple years and there are dozens of public performances along the way, so of course the marketing and audience development is in process that entire time as well. It's an exciting way to work because the art is, in many ways, more about the process and the end result becomes something like an opening night party: important but less so than what's happened to get there.