Running an Agency With 5 Creators: Roles and Thresholds

Five creators is the size where the agency is still one person’s head, and that person is you. You know each creator’s rules, you know the regulars by name, and you can say from memory what last month looked like. That is a real advantage, and it hides the thing that actually limits you here. It is not the number of accounts. It is the number of hours in the evening.

What does a day at five creators actually look like?

Broken into fragments, and the fragments do not fall where you would choose to put them. The work arrives from three directions at once: creators, fans, and the people you pay. Only one of the three keeps to a schedule.

• Morning. Platform statements, overnight messages nobody answered, one creator asking why her payout looks different from last month.
• Midday. Content requests, a price question you have already answered twice, a fan disputing a purchase.
• Afternoon. Next week’s rota, an interview, an invoice, the thing you promised yourself you would write down.
• Evening. You are in the inbox yourself, because this is when fans buy.
• Night. Nothing is staffed, and that is the correct decision.

The evening is the pinch point, and it is structural rather than a matter of discipline. The block that produces almost all the revenue is the block where you are least available, because it is also when creators are awake, filming and messaging you. Every organisational decision at this size is really a decision about who is in the inbox at that hour.

Who does what when the roster is five?

Six jobs exist from your first creator, and at five they are usually held by one or two people. That is workable. What is not workable is leaving them unnamed, because an unnamed job is done last and blamed first.

Job	What it covers	Who holds it at five	First sign it needs its own person
Creator relationship	Rules, limits, expectations, the monthly conversation	You	A creator learns something about her own account from a fan
Shift coverage	Who is in which inbox, and the handover between them	You, plus one or two chatters	You cover a shift yourself because nobody else can
Quality control	Sampling threads and returning corrections	You, badly, in the gaps	You have not opened a transcript in two weeks
Pricing and offers	Tiers, bundles, what is never discounted	You, with each creator	Two chatters quote different prices in one week
Money	Payouts, statements, reconciliation	You	You recompute a payout because you do not trust the first result
Recruiting and onboarding	Sourcing, testing, the first two weeks	You, reactively	You hire only once someone has already left

The pattern in the right-hand column is worth reading on its own: every failure signal is something a person notices from outside, not a number that moves. That is characteristic of this size, and it stops being true later. See running an agency with 20 creators for what replaces it.

Which hours do you cover first, and in what order?

Evenings, seven days a week, before you buy anything else. Coverage is the only spend at this size that buys hours in the block where fans actually buy, and the order in which you buy it matters more than the total.

1. One evening block, seven days a week. That takes two people, not one: nobody works seven evenings. This is the purchase that changes the business.
2. A second person on the peak evening. Same hours, more threads worked properly, and it is where the marginal return is highest.
3. Weekend afternoons. Cheaper to staff than evenings, busier than weekday daytime.
4. Everything else. Including nights. Our data shows conversion collapses between 2am and 6am. Paying a premium to staff the hours that convert worst is an expensive mistake at this size.

Pull your own sales-by-hour per creator first. One account’s peak is not the roster’s. The mechanics of the swap itself sit in shift rotation.

Which tools still hold at five creators, and which stop?

A spreadsheet still holds the numbers and has already stopped holding the rules. That split is the whole answer, and it explains why agencies this size feel simultaneously fine and chaotic.

Job	Spreadsheet at five creators	It breaks when
Revenue per creator	Holds	You need it per creator, per hour and per chatter at once
Payout calculation	Holds, if the base is written down	Exceptions accumulate creator by creator
Rota	Holds	Someone has to cover a no-show the same evening
Voice guide	Holds, one document per creator	Nobody reads it while writing
Handover between shifts	Does not hold	Immediately: a sheet is not open when the message is sent
Quality log	Holds	You have more chatters than you can read

The row that fails first is the handover, and it fails from the first day two shifts run on one account. Everything else can wait. When you do go looking for a tool, judge it on the two jobs a sheet genuinely cannot do (shared live state, and payouts with exceptions), the argument set out in creator management software.

Which four thresholds force a change of organisation?

Four, and not one of them is a revenue figure. Revenue tells you the business is growing; it does not tell you the organisation has stopped fitting.

• The second chatter on one account. Two people writing as one creator, without a voice guide she wrote herself, will contradict each other within a week: a promise ignored, or content re-offered after a refusal. The fix is a real voice guide and a written handover, in place before the second person starts.
• The first evening you cannot cover yourself. The moment the peak block depends on someone else being available, you need a named backup for a no-show. Name the backup now, not at 8pm on the night.
• The week you stop reading transcripts. Quality drift stays invisible in the revenue line for weeks. Once you are no longer reading everything, you need a deliberate weekly sample per chatter, or you are steering on the number that reacts last.
• The payout you recompute because you do not trust it. That is the signal the numbers have outgrown the sheet. Not the size of the payout: the fact that you checked it twice.

Does the fifth creator actually pay for herself?

Only if she does not take hours away from the four already on the roster. Adding an account to a saturated coverage plan does not add revenue. It splits the same hours four ways instead of three, and every account gets worked slightly less well.

Run those figures yourself. Fill this in for one month, before signing:

Line	Your figure
Staffed hours the new account needs, in the peak block	___
Cost of those hours, fully loaded	___
Hours this takes away from existing accounts	___
Your effective share of what she is expected to bring in	___
Net position in month one, and again in month three	___

Two things fall out of it. You need a multiple of the cost back, not the cost, because you keep only a share. And if the third line is not zero, judge the account against what it costs the rest of the roster, not against what it earns alone.

What does better organisation never fix at this size?

What happens inside the conversation. You can name every role, staff every evening and still lose sales in the messages themselves, because the things that move conversion are writing decisions rather than scheduling decisions.

Our data measures two of them plainly. Pitching before the sixth message drops conversion by roughly a third, and the optimum sits after about ten exchanges. And a personal callback placed at the exact moment of the pitch costs points, even though every chatter believes it helps (the most counter-intuitive result in the corpus). Neither is fixed by hiring, and neither is visible on a revenue line for weeks.

Where do you start if you change one thing this month?

Cover the evening block seven days a week, with two people and a written handover. It is the only change at this size that touches revenue, quality and your own availability at once: the money is made in that block, the contradictions between chatters happen at the swap, and until someone other than you covers it, every other improvement happens in the gaps. The rest of the operating rhythm sits in how to run a chatting agency.