I keep seeing the same YouTube thumbnail: "I built a faceless brand for $0." It's never true, and it's never a useful lie, because the interesting number was never zero — it was how the cost is shaped.
I run two content brands more or less on autopilot. This one — an English blog about building with AI — and a second, an Arabic AI-news brand that ships a short every day. Both publish without me touching most steps. So I sat down and added up what they actually cost me in a month, from my own invoices and subscription page, not from a stock spreadsheet.
The total surprised me less than its shape did.
The one line that is almost the whole bill
I pay $100 a month for Claude Max. That is the bill. Everything else rounds against it.
And here's the thing I want to be honest about: that's not a "brand cost." It's the cost of the thing doing the work. It writes drafts, it writes the code that renders the videos, it runs the automations. If I shut both brands down tomorrow I'd keep the subscription, because it's what I use to do everything else too.
Which means the entire premise of "what does the brand cost" is slightly wrong. The expensive part isn't the brand. It's the labour, and the labour is a flat monthly number that does not care how many brands I point it at.
The big line item isn't a cost of running the brands. It's the cost of the worker. And one worker runs both for the same price it runs one.
Everything else, with real numbers
Here is the actual rest of the stack. Where I'm certain of a figure it's here; where I'm not, I say so rather than invent one.
| Tool | What it does | What I pay |
|---|---|---|
| Vercel (Hobby) | Hosts both websites | $0 |
| Domains | The two brand names | ~$11/yr each — about $1.85/mo combined |
| Amazon Polly | Voiceover for the daily videos | $0 (free tier, first 12 months) |
| Email forwarding | Address on each domain | $0 (Namecheap forwarding) |
| Higgsfield | Image/video generation | $0 (on the free trial) |
| Blotato (Starter) | Schedules every post | Paid tier — see below |
| ElevenLabs | Voice for one channel only | Paid tier — see below |
Two of those I won't put a dollar figure on, because I'd be guessing at the current price rather than reading it off an invoice, and pricing pages move faster than blog posts:
- Blotato is on the Starter plan — the thing that lets me schedule every post across both brands and then some. It's the one paid tool that exists because of the brands. It caps me at twenty connected accounts, of which I currently use ten, so it isn't the second brand that pushed me onto it. Check their current pricing for the real number.
- ElevenLabs I pay for, but for a third project's voice — not for either of these two brands. Both of these run their audio on Polly's free tier. I mention it only because it would show up on my card statement and I'd rather over-disclose than have you think Polly is secretly ElevenLabs.
I wrote separately about why the free tool wins the volume work and the paid one wins the hero work — that split is exactly why neither daily brand needs the paid voice.
So what did brand number two cost?
This is the number worth the whole article. I built the second brand after the first, so I can actually answer it.
It cost one domain. About eleven dollars, once, for the year — call it a dollar a month. Everything else it needed already existed: the same AI subscription, the same free host, the same free voice tier, the same scheduler with fifteen empty account slots.
The marginal monthly cost of the second brand, in dollars, rounds to less than the domain. That's the real finding, and it's the one the "$0 brand" videos get backwards. The first brand is where you pay. The second is nearly free, and the fifth would be too. What doesn't scale for free is the only thing that was ever really expensive.
The cost that isn't on any invoice
Time. Obviously. But not the way people mean it.
The automations do the repetitive work — drafting, rendering, scheduling, posting. What they don't do is the thing that keeps the brands from being garbage: reading everything before it ships.
I've been bitten enough to have a rule now. A scheduled task once deployed a half-finished version of this site over a good one, because I told it to run at the wrong moment. Thirty-six posts once sat in a queue with a music-autoplay switch flipped the wrong way that no error would ever have caught. The daily news brand runs on a template, and a template is exactly what Google's own guidelines warn about when it's published without a human in the loop.
So the honest cost of the second brand isn't the domain. It's fifteen minutes a day of a person reading the draft the machine wrote and asking "is any of this wrong, and would I put my name on it." That doesn't show up on a card statement, and it's the line that actually decides whether the whole thing is worth running.
The one bill I'm arguing with
One more, because leaving it out would make this tidier than the truth. I have a subscription to a tool I asked to cancel and am still being charged for while I sort it out — a $44/month line that shouldn't be there. It has nothing to do with either brand. I'm keeping it in the total anyway, because pretending your real monthly spend is only the tools you meant to buy is how people convince themselves side projects are cheaper than they are.
Check your own statement, not your intentions. The gap between the two is usually a tool you forgot you were paying for.
If you want the takeaway without my numbers
Three things, in order of how much they'd change your planning:
- Price the worker, not the brand. The big recurring cost is whatever does the work — the AI subscription, or your own hours. It's flat, and it's shared across everything you point it at.
- Second brand, third brand, tenth — the marginal dollars round to a domain. Free hosting and free tiers don't care how many sites you run until you're much bigger than this.
- Budget the attention, not just the money. The uninvoiced cost — reading everything before it publishes — is the one that decides if any of it was a good idea.
I set out to answer "what do two automated brands cost." The useful answer turned out to be that the question has the wrong subject. They cost one flat subscription, a couple of eleven-dollar domains, and the half hour a day I spend making sure a machine didn't publish something I wouldn't.