Agency, freelancer, or one operator: the same 90 day scope, priced three ways
Take one real scope and price it three ways. The number matters less than what you give up in each version.
the short answer: most comparisons of agency versus freelancer versus consultant argue about which is better in the abstract. that is unanswerable. take one concrete scope instead, price it three ways, and the real differences show up immediately, and they are mostly about who does the work and how fast decisions happen.
The scope
ninety days on a consumer app. validate two acquisition channels, build a creative testing loop producing at least twenty concepts a month, fix the worst funnel drop-off, and hand over documented pass or fail criteria so the system survives the engagement.
Agency
typically $5,000 to $15,000 a month. so roughly $15,000 to $45,000 for the ninety days.
what you get: capacity. several people, coverage when someone is away, and usually a designer, a media buyer and an account manager.
what you give up: the person who sold you the engagement is rarely the person doing the work. your account gets whoever is free, and the seniority you met in the pitch is allocated across a roster. decisions route through an account manager, which adds days.
when it is right: you need breadth across several channels at once and you have the internal capacity to manage a vendor relationship properly.
Freelancer
$100 to $300 an hour, with $175 to $300 typical for seven to fifteen years of experience. at twenty hours a month that is $2,000 to $6,000 a month.
what you get: the actual person, usually cheaply, for a defined piece of work.
what you give up: hourly billing quietly rewards hours rather than outcomes. more importantly, most freelancers are executing a spec rather than deciding what the spec should be, so you still need someone in-house who knows what to ask for.
when it is right: you know exactly what needs doing and need hands to do it.
One operator on retainer
$3,000 to $20,000 a month, so $9,000 to $60,000 across ninety days depending on scope.
what you get: the person doing the work is the person who decided it should be done. no translation layer, no account manager, decisions in the same conversation as the work.
what you give up: capacity and cover. one person has a ceiling, and if they are unavailable the work pauses. that is a real cost, not a footnote, and anyone selling you this model without naming it is overselling.
when it is right: the problem is depth rather than breadth, and you want the system built rather than the tasks done.
The comparison that actually matters
not the monthly number. these three ranges overlap enough that price rarely decides it.
ask instead: how many decisions per week does this arrangement support? growth work generates them constantly. kill this creative, stop this channel, change this onboarding step. an arrangement where those wait three days for an account manager to schedule a call will produce less in ninety days than a cheaper one where they happen on the spot.
and: what remains after the engagement ends? if the answer is a dashboard and a slide, you rented results. if it is a documented testing system your team can run, you bought something.
My bias, stated plainly
i am the third option, so treat this section accordingly.
what i will say against my own model: if you need five channels running simultaneously from next week, an agency genuinely is the right answer and i would tell you that. if you know precisely what needs building and just need it built, a good freelancer is cheaper than me and equally effective.
the case for one operator is narrower than the marketing usually admits. it is best when the problem is that nobody has figured out what should be happening yet, and you want the person figuring that out to also be the person doing it.
if you want your specific scope priced honestly across all three, let’s talk. i will tell you if one of the other two fits better.
ranges compiled from published 2026 market data across marketerhire, growthmentor, robert half and consultant rate surveys. treat them as ranges, not quotes. more detail in what a growth marketer actually costs.