Three ways to run marketing at a 5 to 50 person company, priced against what each one actually delivers in a month, and the question that decides between them.
You have raised money, you have a product, and marketing is now somebody's problem. There are three ways to solve it and most comparisons of them are written by whichever option is paying for the article. Here is the version with the numbers in it.
Hire someone. A competent generalist marketer in a US or UK market costs $90,000 to $140,000 plus equity, so call it $9,000 a month fully loaded before they have written anything. They take four to eight weeks to hire and another six to ramp. They are one person, which means they are good at two of the six things you need and learning the rest on your time.
Retain an agency. Real content and demand agencies start around $6,000 a month and run to $25,000. You get a team, which solves the one-person problem, and you get an account manager, which creates a new one. Output is typically four to eight pieces a month. Most of the retainer pays for coordination rather than production, which is not a scandal, it is just what buying people's time costs.
Run an engine. Software that picks what to publish, writes it, hosts it and sends it. The price range is wide because the category is young. Ours runs $2,500 to $20,000 a month depending on how many channels you turn on and how much of the decision you hand over.
Not price. All three land inside a $6,000 to $12,000 band once you count properly, which means price is a tiebreak and not a decision.
The question is: what is the bottleneck at your company right now?
If the bottleneck is knowing what to do, hire a person. A good marketer with real judgement about your market is worth more than any volume of output, and no software has judgement about a market it has never sold into. Companies at this stage usually know they are here, because their problem sounds like "we do not know what our positioning is" rather than "we do not publish enough".
If the bottleneck is relationships and taste, retain an agency, specifically one that has worked in your category. You are buying their pattern library and their contacts. This is most often true for brand launches and for categories where the buying happens in a small room.
If the bottleneck is throughput you want an engine, and this is the case nobody says out loud because it sounds unromantic. You know what to say. You know who to say it to. You are producing two pieces a month when the plan called for twelve, and the reason is not strategy, it is that somebody has to sit down and write them.
A hire fails when the company cannot brief. One marketer inside a company with no clear position spends six months doing discovery and then leaves. If you cannot write down who you sell to in a sentence, hiring will not produce that sentence for you.
An agency fails on the second quarter. The first is genuinely good because you get the senior people. Then you are handed to the team who actually staff the account, and the work regresses to their house style rather than yours. This is not malice, it is how the economics of an agency work.
An engine fails when it is asked to have taste. Software can pick a search worth writing for and produce a competent piece against a brand it has been given. It cannot decide your positioning, it cannot read a room, and it will confidently produce twelve pieces of nothing if the inputs are wrong. That is the honest boundary and it is the reason we gate our own inputs before we generate anything.
| | Hire | Agency | Engine | |---|---|---|---| | Pieces produced | 4 to 10 | 4 to 8 | 12 to 60 | | Channels covered | 2 to 3 | 2 to 4 | All configured | | Turnaround on a change | Hours | 3 to 10 days | Minutes | | Judgement about your market | High | High | None | | Cost of the eleventh piece | Overtime | A new line item | Zero |
The last row is the one that changes the arithmetic. People and agencies price per unit of attention, so doubling output roughly doubles cost. Software prices per capability, so the marginal piece is close to free. That is the entire reason the category exists, and it is also why "which is cheaper" is the wrong comparison: they have different shapes, not different prices.
One person who owns positioning and judgement, and an engine that owns production. The person writes the brief and reads the output. The engine turns one brief into the twelve assets nobody had time to make.
That is roughly $11,500 a month at our entry tier, produces more than an agency at twice the price, and keeps the judgement inside the company, which is where it belongs and where it compounds.
Whatever you are leaning toward, buy the smallest version first and check one thing: does the output survive contact with someone who knows your customer? Give the hire, the agency or the engine the same brief. Read the three results side by side. The one that gets your customer right is the one to buy, and you will know within a week rather than a quarter.
If you want to see what an engine's read of your brand looks like before you decide anything, we run ours on any site for free and send you what it found. No call, no deck.
Persistent brand memory. Self-correcting generation. Built for DTC operators who can't afford generic copy.
Book a discovery call →