Publishing a price list before the call is a harder decision than it sounds. The usual advice is to keep the number for the call, because the call is where you find out the budget. I stopped doing that, and the reason is mechanical: if the price is hidden, the first fifteen minutes are spent on the price instead of on the bottleneck, and the bottleneck is the only thing that decides whether the number is expensive.
So: Sprint $900 one-time, Engine $1,900 per month, Full build $2,900 one-time. No hourly rate, no discovery fee, no "let's scope it on a call".
What each package actually buys
Sprint — $900 one-time. One channel or one path, built and live: the capture point, the sequence, the tracking, the kill rule. This is the right first purchase when something is already being sent and nobody can say whether it works. One path, instrumented, with a number that can be read on Friday.
Engine — $1,900 / month. The continuing block: build, measure, kill the losers, scale the one that survived. Fourteen days' notice to cancel. A retainer you cannot leave is not a service, it is a lease.
Full build — $2,900 one-time. The whole system shipped and handed over running: offer, pages, outbound, measurement, documented handover. Chosen when the assets have to survive without the operator in the room.
Fixed packages exist because hourly billing charges you for the time it takes to find out what the problem was. The unit of work here is an artifact with an openable link: a live page, a published sequence, a measured event, a documented decision.
The comparison you are already making
Market ranges below are the usual advertised ones, not a claim about any specific provider.
| Option | Typical cost | What that money actually buys |
|---|---|---|
| Fixed packages | $900 sprint · $1,900/mo · $2,900 build | Prices published up front, an artifact list as the deliverable, no account-manager layer, assets stay in your accounts |
| Freelancer on hours | $40–$120/hour, uncapped | Cheap per hour, expensive per outcome: the same work is billed by the clock, and the useful question — did anything ship — is not on the invoice |
| Agency retainer | $3,000–$10,000/month + ad spend | A team and a reporting rhythm. Worth it when you need several specialists at once; heavy when the real problem is one broken path and no measurement |
| Full-time hire | $90,000–$150,000/year | Availability. Rational once the work is continuous and the offer is proven; the wrong first move when the bottleneck has not been identified yet |
A marketing engineer is not cheaper than a freelancer and not cheaper than a junior hire. It is priced against the cost of a month that ships nothing.
Two questions beat any rate card
Who owns the asset when the engagement ends? If the answer is "our platform", you are renting the thing you paid to build. Every asset should be in your accounts, in your repo, on your domain.
What exactly can you open to check that a month happened? Not a slide. A URL: the live page, the published sequence, the dashboard event, the written decision with a date on it. If a proposal survives neither question, its price is irrelevant.
What moves the number up or down
- Mechanical bottleneck, settled offer → the cheapest kind of work. One path, built and instrumented. Sprint territory.
- Unsettled positioning → costs more on purpose, because wiring a funnel nobody wants to enter just makes disappointment arrive faster. That work starts on the page and the offer, not on the machinery.
- No tracking at all → one event and one analytics property are part of the first deliverable; otherwise every later decision is an opinion.
- Several channels at once → not a discount, a sequencing problem. Run parallel tests without sample-size discipline and you get a confident answer to the wrong question.
- Ad spend and tool subscriptions → never included. The engagement buys work, not budget; a service that pays for your media is a service that needs your media to keep paying.
The number is the easy part
The expensive part is a month where the pipeline produced a report and no verdict. Three prices are published, the scope is written out, and the deliverable is a list of artifacts you can open — the full list and the honest comparison table live here: marketing engineer cost.
If the bottleneck is already identified and you want the path built and measured: what that looks like. If you would rather check the work before the price, the scope of work and the proof page are public too: scope · proof · pricing.
Ask for the artifact list before you pay anyone. Including me.
Top comments (0)