All posts
4 min readFounder essays · Operations

Fire your agency, keep the results

The pitch sounds reckless until you look at what the retainer was actually buying. Most of it stopped being a human job, and nobody updated the price.

"Fire your agency" is the kind of line that sounds like marketing bravado until you actually itemize what the agency was doing. We put it on our homepage on purpose, fire your agency, keep the results, because the provocative part isn't the firing. It's the keep the results. The claim only works if the output survives the change, and whether it does comes down to one question: how much of what you were paying for was genuinely human, and how much was just expensive by habit?

What the retainer actually bought

Pull apart a typical social retainer, the $3,000 to $12,000 a month a competent agency charges, and it resolves into four jobs.

There's strategy: what to say, to whom, and why. There's production: turning that into actual carousels and captions. There's translation: adapting each piece so it lands natively on each platform. And there's consistency: keeping it all on-brand and shipping on schedule, week after week.

For most of the last decade, all four were human work, and the retainer was a fair price for four humans' worth of effort. The thing that's changed isn't the value of the output. It's that three of those four jobs stopped requiring a person.

The retainer didn't get overpriced. Three of the four things it bought just stopped being human jobs, and nobody re-priced it.

The three jobs that left

Production is the most obvious. Once your brand is captured as a structured kit, your real colors, your fonts pulled from your own CSS, your voice, generating an on-brand carousel is a pipeline step, not an afternoon in a design tool. flypost.ai does it from a pasted URL in about a minute for the first drafts, and on demand after that.

Translation went the same way. Re-rendering one idea as a LinkedIn carousel, a tight X post, a portrait Instagram set, and a native Reddit post used to be four discrete tasks a human did four times. It's now per-platform captions, hashtags, and image dimensions produced in a single pass.

Consistency, the quiet, grinding job of not drifting off-brand and not repeating yourself, turned out to be the most automatable of all, because it's fundamentally a checking problem. Every candidate post bound to the brand kit, every angle embedded and compared against your full history, anything within 0.85 similarity of a past post dropped before it ships. A human doing this by hand is slower and worse at it than the check is.

That's three of the four jobs, the three that made up most of the hours and most of the bill, now done by software.

The one job that stays

The fourth job, strategy and taste, does not leave, and any honest version of this argument has to say so. Deciding what's worth posting, whether a draft actually sounds like you, when to react to something happening in your market, that's judgment, and judgment is exactly the part you shouldn't automate.

But here's the thing founders discover when they look closely: that part was never what the retainer was really charging for, and more to the point, it was always partly yours. Nobody understands your business's point of view better than you do. The agency's strategist was often translating your judgment, not replacing it. When you keep the taste in-house, where it already half-lived, and hand off the production, translation, and consistency, you haven't lost the results. You've kept the only input that was ever truly load-bearing and shed the labor around it.

The math, without the spin

Lay the two paths side by side. The agency: $3,000–12,000 a month, two to four weeks before anything ships, eight to twelve carousels, two or three platforms, locked into a six-to-twelve-month contract. The autopilot: from $49 a month, set up in under a minute, a month's worth of carousels with no per-post fee, five platforms including Reddit, month-to-month, cancel anytime.

The agency figures are typical 2026 market ranges and yours will vary. But the gap isn't a rounding error you can argue away with "you get what you pay for." It's two orders of magnitude, for output that, on the dimensions a founder actually measures, on-brand, native per platform, consistent, scheduled, is comparable. You're not paying ten times more for ten times the quality. You're paying ten times more for labor that no longer has to be bought by the hour.

The honest caveat

There's a version of "fire your agency" that's genuinely wrong, and it's worth naming so this doesn't read as zealotry. If you're running big paid campaigns, need real creative direction on a complex brand, or are staffing a launch that touches a dozen channels, a good agency or a senior in-house hire earns the money. Automation replaces production and consistency; it does not replace a strategist on a hard account.

But most founders, solo marketers, and small teams aren't in that situation. They're under-posting because production is a weekly tax they can't afford, and the only solution anyone offered them came with a comma in the contract. For them, "fire your agency, keep the results" isn't bravado. It's just arithmetic that finally caught up to what the tools can do, the production, the translation, and the consistency handed to a pipeline, and the taste, the part that was always the point, kept exactly where it belongs.

Keep reading