Why the initial phase weighs more than it looks
The launch phase of a marketing or automation project involves diagnostic and correction tasks that don't repeat later, unlike the regular monthly follow-up work. That's why it's worth treating it as a separate deliverable with its own cost, rather than quietly folding it into the standard monthly fee.
Before any campaign or workflow runs properly, there's always a phase of getting the house in order: checking whether the website works, whether pixels are installed, whether ad accounts exist, whether Analytics is configured. This diagnostic and correction work doesn't repeat later, unlike the monthly follow-up tasks that come after.
I describe this moment as a checklist: I look at the landing page, the tracking tools, the audiences already built, and I produce a recommendations document before launching any campaign. When that review reveals substantial correction work, that's when I charge a setup fee, which can typically go up to double what the client will pay in the following months. I don't apply this to every client; it's the criterion I use when the upfront effort clearly justifies a separate charge.
What stays with the client versus ongoing attention
The initial setup work creates something that remains with the client even if the collaboration ends, while the monthly fee pays for continuous attention afterward. That difference in nature between building something and watching over it justifies separating the two in a commercial proposal instead of blending them into one undifferentiated invoice.
If I fix the landing page, build the campaigns, install pixels and organize Analytics, that stays done, with or without me the following month. The monthly fee, on the other hand, covers ongoing follow-up: adjusting campaigns, reviewing metrics, reacting to changes in the market or on the platforms themselves. If I charge the same amount for both moments, I'm effectively offering the first one as a bonus, hoping to recover its value across many months of retainer, which only works if the client stays long enough to amortize that initial investment.
I deliver faster than before. How do I defend my service price?
How to organise a project's first month without promising to fix everything
Growing an online business: what I learned about margin, team, and AI
To design execution and handle exceptions, explore the business process automation guide.
The risk of unvalidated businesses
Most businesses that start working with digital marketing without a validated product or sales process end up failing, regardless of the quality of the technical work delivered. That changes the calculation of how much it's worth investing upfront without charging for it, especially during the first months of any new partnership.
I'm blunt about this: most unvalidated businesses that start working in digital will fail, often within just a few months of work. If I put a heavy load of unbilled work into the first month without any guarantee of contractual retention, the risk of losing that time investment is real and recurring. This directly shapes how I price the setup of any new engagement.
Negotiating when the client won't pay for setup
If a client resists paying an upfront fee, the alternative I use is trading that charge for a longer commitment written into the contract itself. It's an explicit trade of risk for risk between both parties, giving up one form of protection in exchange for another.
The way I put it is simple: if we sign a 12-month contract, I won't charge the initial setup, and I consider that an extremely valid and extremely fair trade-off. If the client wants the security of not paying upfront, I need the security of knowing I'll have enough time to recover that investment through the monthly fee. What I avoid accepting is both at once: a heavy unbilled setup combined with no guaranteed length of contract whatsoever.
Don't try to solve everything in week one
One mistake I acknowledge making is trying to fix everything immediately at the start of a project, which creates a pace that later turns out to be impossible to sustain. The client quickly gets used to that initial intensity and clearly notices the difference once it settles into the normal maintenance rhythm.
I tell my students: don't try to solve every problem in week one, this was probably one of the biggest mistakes I made. Dedicating excessive hours to a client at the start, especially when you still have few clients on your roster, feels like genuine generosity, but it creates an expectation that becomes unsustainable as the business grows and more clients arrive. I compare it to a romantic relationship: at the start there are many gestures and plenty of attention, and over time it's normal for the pace to settle naturally. The problem arises when the client experiences that shift as a loss of importance rather than understanding it as a healthy normalization of the working relationship.
Meetings with a purpose, not by habit
Meetings consume valuable execution time and I only schedule them when there's something concrete to decide, never simply because it was set to meet "every Monday" without any other reason. Setting a realistic follow-up cadence protects the time of whoever is actually doing the technical work day to day.
I suggest dividing the project into phases with clear reassessment points, rather than filling the calendar just because it looks more professional to the client's eyes. Early in a project it may make sense to meet more often during the week; after a few months, a single monthly check-in might already be more than enough. The criterion I use is always whether there's a concrete, defined objective for that particular conversation.
Applying this to automation and AI projects
This reasoning about setup versus retainer came from a class on onboarding digital marketing clients, but it applies naturally to any project where the initial configuration phase is clearly heavier than the regular operation that follows, including automation or AI tool implementations being built from scratch for a client company.
If you're building workflows, cleaning databases or configuring integrations that keep running autonomously once delivered, it's worth treating that setup phase as its own deliverable and deciding, case by case, whether to charge for it separately or ask for a longer commitment in exchange for waiving that upfront charge. This is a reasonable extension of the commercial principle I explain in the context of digital marketing, not a fixed description of what I do in every automation project.
For those working in automation and AI consulting, this is one of the topics we explore in the Laboratório da IA mentorships: how to structure the commercial proposal so that the initial technical effort doesn't fall entirely on the service provider.
Source note
This article is based on Sexta Ímpar #89, a class on the onboarding process for digital marketing clients. Around the 17 and 27 minute marks, I talk about setup fees, contract trade-offs, and the mistake of trying to solve everything in the first week. Full video here: https://www.youtube.com/watch?v=LEAed-8P2is.