All posts

Positioning

The operator playbook: running 50 brands without losing your sanity

Why Buffer's per-account pricing breaks at 10 brands, what we built instead, and the unfair flat-rate math that makes it work.

May 25, 2026·6 min read

Most social media schedulers price per social account or per seat. That model fits a team of marketers running a single brand. It scales poorly when a single person runs many brands — a profile we see in indie hackers, portfolio operators, holding companies, and small studios.

Postcrow is designed for that profile. This article explains the design choices: why we chose flat-rate pricing, how the product reflects multi-project use from the ground up, and which workflows it's optimised for.

The pricing problem

Per-account pricing scales linearly with account count. For a team managing one brand, the cost is bounded and predictable. For an operator managing ten brands, the same pricing structure means ten times the cost — even though their support burden and usage pattern haven't scaled the same way.

Postcrow uses flat-rate pricing instead. €49/month for the Agency plan covers unlimited projects, unlimited social accounts, and unlimited AI captions. The marginal cost of an additional project in our system is small enough that we can absorb it without changing the price.

BrandsPostcrow AgencyPer brand / month
1€49€49
10€49€4.90
50€49€0.98
100€49€0.49

For comparison points, public pricing pages of Buffer, Hootsuite, and Later are easy to find. The point isn't that they are wrong for their market — they're built around different assumptions. The point is that a portfolio operator's usage doesn't fit those assumptions cleanly.

What the product reflects

Flat pricing is necessary but not sufficient. Running 50 projects in a tool that wasn't designed for it is its own kind of slow. Four product primitives matter:

1. Project switching

At 50 projects, a dropdown is the slowest part of the workflow. Postcrow has a Cmd+K command palette that fuzzy-finds any project in under a second. You can use it from any page in the app, including the keyboard shortcut equivalents on Linux and Windows.

2. Brand voice per project

With many brands, you can't remember each brand's tone when you draft a post. Postcrow lets each project store a trained brand voice — a paragraph distilled from 3–10 sample posts — that gets auto-injected into any AI compose call for that project. The details are covered in Brand voice training.

3. Bulk operations

Some workflows scale better as batches than as single-post UIs. Postcrow ships a CSV importer (up to 1000 rows per import), project cloning (timezone + brand voice copied in one click), and a bulk-schedule primitive available both in the UI and from the MCP server. Bulk operations covers each in detail.

4. A terminal interface

Many portfolio operators are technical and spend their day in an editor or terminal. Postcrow ships a Model Context Protocol server so the full feature surface is reachable from Claude Code, Cursor, or any MCP- compatible client. You can list, draft, schedule, recycle, and pull analytics without opening a browser. See the MCP server article for details.

What Postcrow is not

Postcrow is not built for large teams running a single brand. Team approval flows, role- based permissions, and internal review queues aren't in scope today. If your bottle neck is internal coordination on one brand, established tools serve that case well.

We focus on a different bottleneck: the moment one person tries to run more brands than the per-account economics of those tools allow.

How to start

Sign up, create one project per brand or clone an existing one, connect at least one social account, train each project's voice, and either compose manually or enable auto-recycle so the queue stays alive without daily attention. If you already use Claude Code, generate a Postcrow API key and add the MCP server.

The 14-day Pro trial is free without a card. Agency is €49/month, flat.