What came before: We grew our new organization, Luna Burn, from two people to 100+ in 5 weeks. Rapid growth teaches lessons that slow growth doesn’t. We hit organizational scale problems almost instantly.

We were barely a week old and already bumping up against structural problems of scale. We’d seen how a single chat couldn’t work for coordinating too many people. Rapid scale forced structure and communication systems.

Organizational scale forces structure, but scaling a customer base forces other things as well.

Scale Makes Anything Possible ... Especially Chaos

The event land could handle 1,500 people. Past events had been that big. But how big should this event be? Insurance and medical services get tricky the more people you have.

More people means we need more of what we do already. We’ll pay more invoices, field more queries, and order more portopotties. But the the crowd quietly brings New Stuff to handle. Not just the need to manage a larger stuff, but genuine New Stuff that only happens with size.

Some things in life happen rarely. But with more attendees, more customers, and more employees, those rare happenings become annoyingly frequent.

One in a hundred customers drops their Turnip Twaddler on a hard floor and it breaks. With 5 customers, you rarely (if ever) get breakage. Everyone treats their Turnip Twaddler with kid gloves. But once you have 5,000 customers, you can expect 50 exchanges.

The customer thinks this is rare. After all, they have never before dropped a Turnip Twaddler. This is a once-in-a-lifetime disaster! To you, however, it happens daily. You might event keep an extra box of Turnip Twaddlers handy to give your customers.

As you grow bigger, you must plan for all those previously rare things that now happen regularly. Customers aren’t getting more careless; you just have so many of them that some rare thing is always happening.

Chaos Reigns Outside

The problem isn’t just clumsy customers. Rare events also become more common with other outside stakeholders. You have more vendors and more funders. Vendors need to be coordinated logistically and in terms of quality. The rise of rare events happens here, too. When you receive 100 parts from a vendor, quality control is relatively easy—you can inspect each one if necessary.

And as for funders... funders need to be managed. Some investors or creditors need regular reporting. Some have a board seat, or an advisory position, or retain the right to examine the books. The more of them you have, the more of their quirks you have to manage, not to mention any political battles that they’re fighting among themselves. Compliance, oversight, and investor management becomes a full-time job.

... and Chaos Reigns Inside

There’s an even worse hidden gotcha (as if that weren’t enough):

Rare events get more common inside the organization. Depending where they happen, those events can do lasting good or inflict lasting doom.

Organizations are amplifiers. If you hire that rare front-line employee who doesn’t care about customers, the worst case is that they alienate a few customers. No biggie.

If you happen to hire that rare vice president of customer service who doesn’t care about customers, the not-caring scales. They create systems, hire employees, and set incentives, all based on not-caring. The more they do, the harder it is to undo. Soon, not-caring is set in stone. It long outlives them, and you end up with an organization where no one treats customers well. Undoing that damage is difficult, if not impossible.

Scale brings rare chaos, and increases the chances of ending up with that unqualified vice president wreaking damage that’s magnified by their role.

Some Rare Things Also Magnify the Damage

We’ve seen so far that size brings rare events, inside and out. Those rare events, which weren’t a consideration before, now need to be dealt with. But while scale increases the kinds of events you need to deal with, it also means that all events can do much more damage.

One of the rare things that happens at an event like a regional Burn is someone deciding to leave and abandon their stuff. Campmates and event staff are left to clean up the mess. Obnoxious, sure. But the burden to the event is small.

Other rare things affect everyone. A few years ago, I was helping with a regional Burn when we were hit by sudden heavy rains. With 20 attendees, we could have tossed our stuff in our cars and left.

But we didn’t have 20 people. We had 1,500 people, each with camping gear and artwork. We had to get them all out of the camping area, down to the parking lot, and onto the small back roads leading to the event. That takes a dozen traffic directors, a logistics plan, and a very high tolerance for upset attendees, who all want to jump to the head of the line to get home.

You can’t evacuate that many people quickly. Just getting cars out of the parking lot takes hours. At that size, you need plans for rare events, and you need contingency plans for extremely rare events if there’s the chance of great damage.

The Roadmap to Scale

While it doesn’t make scale any easier, knowing these “gotchas” of scale can help you plan for tomorrow’s growth phase before you get there. For us, the growth happened so fast that we were always thinking a week ahead ... because growing at roughly 40% each week, every seven days we hit a new gotcha.

If you would like a summary of the ways in which scale changes your business, you can download a copy at https://steverrobbins.com/scaleroadmap2

Subscribe now

The stage: We grew our new organization, Luna Burn, from two people to 100+ in 5 weeks. Rapid growth teaches lessons that slow growth doesn’t.

Scaling is a bitch. Everyone says so. But why? Large corporations are notoriously big and slow, while small ones are nimble and creative. The reasons are structure and, surprisingly, statistics.

I’ve worked with companies from two to 2,000 people.

With two people, we were a rocket ship. When we needed a website, it took 15 minutes, and that was with one hand tied behind my back.

In a 2,000-person organization, a one-page website took six weeks. Six Weeks. I agonized. “How could this possibly take this long?”

The villain is scale. Scale makes jobs narrow. It requires new systems, and new communication. More bad things happen, and some of them are worse.

And when you’re growing as fast as Luna was growing, the problems hit fast and hard.

Everything, Everywhere, All at Once

Luna started with a single Signal chat. Everyone who wanted to be part of the effort was part of the chat, and it was growing by the hour.

Everyone was discussing everything:

• What did we want our culture to be?

• How were we going to get the money to run an event ourselves?

• What departments were going to exist?

• Would it be possible for us to rent the land?

This is the “big bang” moment in a startup. Everyone’s a generalist.

Organizations Grow, Jobs Shrink

Imagine that say you start with one person, Europa, handling both customer support and accounting. She handles everything. She knows every customer’s status. She handles both customer support and accounting, in one convenient person.

Then sales takes off. You’re up to 500 customers, and even Europa can’t track all that by herself. You hire more help—three more customer support people and two more accountants. Europa’s former single job of support+accounting is now a bundle of smaller jobs: support people who each handle a small part of the total support that needs to be done, and accountants who each handle several customer accounts.

The original doer, Europa, either becomes a single customer support rep or ends up managing support reps—a very different skill set than being a support rep herself.

As the business grows, the overall job grows. The bigger job requires more people. Each person gets a part of the overall job. This means that individual jobs are narrowing in scope, while increasing in depth.

Scale Demands New Systems: Enter Managers

As work gets bigger and jobs get smaller, you need to coordinate all those small jobs so they end up doing the scaled-up version of the big job they once were.

This is what a manager does. Europa, once the do-er, is now the only person who knows everything that needs to happen. So like it or not, she ends up needing to coordinate all of the narrow-but-deep new jobs.

Coordination Takes Communication

Narrow-but-deep jobs need to coordinate among themselves. When everyone is in one room discussing everything, it works. A customer service request comes in, and the support person turns to the accountant and asks, “Is this customer paid up on their support contract?” But the customer base grows. Eventually, there could be fifteen support reps and four accountants handling customers.

But when one of fifteen support reps gets a customer request, they need to make sure they’re not duplicating the work of some other rep. If a customer calls five times asking for a refund and gets five different customer service reps, you don’t want all five of them to separately issue a refund. This takes coordination within a function so everyone acts together.

One Week in and We Already Had Silos!

We hit scale limits almost immediately. As our signal chat grew, having a dozen conversations in one channel was driving us crazy. So we moved to Discord (another communication platform) and created channels for different topics. We had one for finance, one for art, one for graphic design, and so on. That way, each topic had one channel where they could have a coherent, single-themed conversation. We’d solved the problem of communicating within an area.

And that led to a whole new problem: the same issue would pop up in multiple channels, with multiple groups. At one point, three separate channels were discussing parking policies…and drawing opposite conclusions.

Our silos needed to coordinate between each other. Remember Europa’s fifteen customer support reps? They need to ask the four accountants if a caller has paid for support. What once happened in Europa’s head—checking with accounting—now needs to happen between two departments.

Yes, scale drives structural changes ... but wait, there’s more

To recap: when we were one chat with a few people, we didn’t need structure. Growth required separate groups to take on different aspects of the work. Within each group, jobs began to shrink. People needed to coordinate with each other. And as each group’s attention shifted towards working within the group, they lost touch with the other groups, driving the need for communication between what were fast becoming departments.

We put systems in place as fast as we could. Everything was ad hoc, since we had only four months before the event, we needed speed more then elegance. Nevertheless, regardless of our timeline, without managing the communication channels, there was no way we would pull this off.

But then statistics, glorious statistics, threw us an unexpected curveball. We’ll pick that up in the next installment.

If you would like a summary of the structural changes scale produces, you can download a copy at https://steverrobbins.com/scaleroadmap1

The conflict wants a yes/no answer. There isn’t one, because it’s the wrong question. What really matters is the meaning that “I’m a writer” has. That is what’s provoking all the feelings.

“Writer” Means Something. Literally.

To know if you can say “I am a writer who uses AI” or “I am a coder who uses AI” or “I am a whatever who uses AI,” let’s review how words work.

Words mean things. Writer, coder, and painter are agent nouns. That means they are doing the action. Here, those nouns are one-who-writes, one-who-codes, and one-who-paints. That’s the claim the word is making.

Does using AI make you less of a writer? It depends. Writing isn’t one skill; it’s many. It has many subskills: creating ideas, developing logic, outlining, researching, wordsmithing (aka “writing”), editing, and proofreading.

When one person does everything, we call them a “writer.” When we give an LLM all those roles with a prompt like “write a science fiction story about a mouse that can talk,” then by definition, the LLM is all of those doers. If all you’re doing is giving a prompt, then you’re not writing; you’re commissioning writing from an LLM. The LLM is the writer.

Superman, X-Men, and Archie show us the way

Comic books give us a way to talk about this. Creating a comic takes many skills: writing the story, penciling the layouts, inking the final linework, coloring the pages, and lettering the speech bubbles. Sometimes one person does everything. But not always.

In companies like DC, Marvel, or Archie Comics, those are separate roles, done by separate people. Each is considered a distinct, valuable skill set.

The word for someone who does everything themselves is “cartoonist.” It means they take all the roles. Otherwise, they name their role. “I’m the writer of X-Men.” Or “I’m the inker for Superman.” Or “I’m the colorist for Elfquest.”

Writing Isn’t as Clear-Cut

We’re sloppier about the word “writer.” But perhaps we can resolve the holy war of job titles by taking a page from our comic book brethren.

I’m a writer. I choose the words and the sentences. I have all the skills—I’ve done every task from idea to published work—but often, it’s a team effort. Having a separate editor (mine are almost always named Emily for some reason) improves my work. I’m still the writer, and Emily is the editor.

The Anatomy of Writing

There’s more to producing a written work than just writing and editing. There’s ideation, outlining, research, developmental editing, writing the words, copyediting, proofreading, and coordinating all of the above.

What if you use an LLM to come up with ideas, develop the outline, do the research, and do all the editing, while you write the words? I think that makes you a writer. But you’re not the idea creator (who I’ll call the “ideator”), you’re not the outliner, you’re not the researcher, you’re not the editor, and you’re not a proofreader. You are one-who-writes. It would be misleading to take credit for the other phases.

There’s an easy test to find out if you inhabit a particular role. Just imagine someone asking you a question about that role in an interview.

• How did you come up with that idea? If you had inspiration other than asking an LLM, then you were the ideator.

• Why did you write this article chronologically rather than as an expository essay? If you had a logic to the article outline, then you were the outliner. Otherwise it was the LLM.

• Why didn’t you write in a more academic style? If you can say why you made that decision, then you were the writer. Otherwise it was the LLM.

AI Nutrition Labels Can Heal the Divide

To help quell the holy wars over whether someone is a real writer, I propose we use a label, just as we do with food. An “AI Nutrition Label” can address the separation of roles head-on. It will tell skeptical readers where we applied our own expertise and where we used an LLM.

Here’s the AI Nutrition Label for this very article. Claude helped me frame the idea, learn the specific names for the roles in the comic book industry, and do an editing pass before publication.

─────────────────────────────────
🏷️  AI NUTRITION FACTS
    Serving size: 1 article
─────────────────────────────────
                    HUMAN    AI
Ideating              ✓       ✓
Outlining             ✓       ·
Researching           ·       ✓
Developmental editing ✓       ·
Writing               ✓       ·
Copyediting           ✓       ✓
Proofreading          ·       ✓
Coordinating          ✓       ·
─────────────────────────────────
Stever Robbins is the writer of
this article. Claude assisted.
─────────────────────────────────

Writing a label like this is also a tool for self-insight. As I wrote the label above, I definitely felt a twinge of vulnerability. “Will people think less of me because I used Claude to do some research? Is the title ‘writer’ justified if I brainstormed with Claude to come up with article ideas?”

That vulnerability teaches me something. It’s my mind’s signal that I want credit for doing all the work. It might take longer, but that’s what makes it my work. Not everything is just about the end result; sometimes the process matters.

After reading the label above and noticing my feelings about it, I’ve come to realize that I want to own as much of the process as possible. Outsourcing the copyediting and proofreading makes sense—editing my own writing is hard to do because I’m too close to it. But everything else? That’s mine.

You Can Be What You Want—Just Do the Work

Being a good writer from idea to finished work is hard. But it’s hard because the most important parts of writing are developing and structuring the message the writing must convey. It’s word choice. It’s logic. It’s understanding your audience, what they know, and what they don’t. It’s crafting a piece of communication that reaches this audience with a message designed to change the audience in this way.

Use an AI nutrition label so others know—and you know—your actual contribution to the writing you’ve created. If there are many checks in the “AI” column, you can decide if you want to develop those skills or continue to outsource them. You can learn to do all the phases yourself, then outsource the ones you don’t like.

I guarantee you’ll love yourself more.

Writing is not the skill of putting sentences on a page. It’s the skill of putting the right sentences on a page. Doing the work yourself doesn’t just demonstrate the quality of your thinking—it improves it.

Then when you step in front of an audience, sit down for an interview, or discuss your writing one-on-one with an employer or prospect, you’ll be able to answer any question in a way that shows strength, competence, and rigor.

One of my favorite business school memories was of a case protagonist who spoke to my entrepreneurial management class. His company did something unfathomably unsexy. Maybe they made aglets (the plastic tip of shoelaces) or something like that.

But the market was so unsexy that no one else was in it. They had excellent profit margins, great work-life balance for all employees, and he pocketed the profits every year.

For an hour and 20 minutes, a bunch of 26-year-old HBS students discussed how he should reinvest profits, borrow money for leverage, expand his market presence, adopt new product lines, roll-up his industry, divest underperforming product lines, yada yada yada.

When it was finally his turn, he got up and stood there quietly. Then he said:

"I take home $4 million after tax and work about 20 hours a week. What, my friends, is so wrong with that picture that I need the stress, complexity, and risk of all this growth.

"Do any of you make that kind of money? Do you have any concept of whether that's enough to have the life you want?

"Come talk to me once you're at that level, and I'll listen seriously to your recommendations." That comment has stayed with me for 36 years as our entrepreneurial-industrial complex has normalized talk of billion-dollar valuations, and neglected value creation at smaller scales.