Marketing automation has a pricing problem: the platform subscription — the number everyone quotes — is usually the smallest part of what you'll spend. The honest answer to "what does it cost?" has three parts: the platform, the build, and the upkeep. This post puts real numbers on the first, is straight with you about the second, and gives you a way to calculate whether any of it pays.
Platform prices below were checked against the vendors' own pricing pages on August 15, 2026. They change often — treat the links as the source of truth.
Part 1: The platform (the number everyone quotes)
For the automation engine itself — the thing that watches for a new lead and fires the follow-up — current pricing:
| Platform | Free tier | Paid entry | Notes |
|---|---|---|---|
| Zapier | 100 tasks/mo | Professional from ~$20/mo (annual) | Scales to ~$130+/mo at 5,000 tasks |
| Make | Yes | $9–29/mo | Cheapest per-operation at moderate volume |
| n8n | Self-host free | Cloud from €20/mo | Self-hosted: roughly a $10–20/mo VPS, unlimited runs |
A typical small business running lead capture, a welcome sequence, and CRM sync lands between $20 and $100/month in platform costs. High-volume businesses hit the interesting divergence: at ~100,000 operations a month, Zapier can run $500–1,000/month while self-hosted n8n still costs a VPS. We've broken down when each platform makes sense in our Zapier vs Make vs n8n comparison.
Add the email side — every serious sequence needs a sending platform — and most small businesses add another $0–100/month depending on list size. Our email tools comparison covers those numbers.
Part 2: The build (the number nobody quotes)
Here's the part the platform pricing pages skip: someone has to design and build the workflows, and this is where most of the real cost lives — whether you pay for it in money or in your own evenings.
We won't quote you fake industry averages — the "$3,000–$15,000 typical setup" figures you'll find on Google trace to content farms, not research. What we can tell you is what determines the cost, because it's the same list we use to scope real projects:
- How many systems are involved. Form → email sequence is an afternoon. Form → CRM → enrichment → routing → sequence → reporting is a project. Each connection multiplies the testing surface.
- Whether your process is actually defined. Automation encodes decisions: who counts as qualified, when to follow up, what happens on no-reply. If those answers don't exist yet, defining them is the project — the workflow build afterwards is quick.
- Data state. Syncing a CRM full of duplicates and dead contacts automates the mess. Cleanup is unglamorous, often necessary, and belongs in the budget.
- Failure handling. The difference between a demo and a production workflow is what happens when the API times out or a lead submits twice. Hobby builds skip this; that's why they quietly break within months.
The DIY math: multiply your realistic build hours by what your time is worth per hour, and compare it against a fixed-scope quote. Sometimes DIY genuinely wins — usually for the first, simplest workflow. It's the third and fourth ones, tangled with the first two, where self-built systems start costing more than they save.
Want a scoped number instead of a range? Book a free 30-min strategy call →
Part 3: The upkeep (the number that decides long-term ROI)
Automations are software, and software needs maintenance: apps change their APIs, fields get renamed, a workflow that ran for a year fails silently on a Tuesday. Budget a few hours a quarter — yours or a partner's — or accept that "we automated that" will slowly become "we thought we automated that."
This is also the fairest criticism of automation for very small operations: if the workflow saves two hours a month, maintenance can eat the entire return. Which brings us to the deciding step.
Do the math before anyone's pitch — including ours
For context on adoption: half of US small-business workers now use AI tools at work, but only 6% use them for genuine workflow automation (US Chamber of Commerce Foundation / Ipsos, May 2026) — the gap between "we use tools" and "our processes run themselves" is where both the opportunity and the wasted spend live.
The way to know which side of that line you're on is arithmetic, not persuasion: hours currently spent, times hourly cost, times the share that's realistically automatable, minus tools, against setup. Our free automation ROI calculator does exactly that calculation, with the assumptions visible. If the payback period it shows you is measured in years, don't automate that process — and be suspicious of anyone who says otherwise.
If it's measured in weeks, that's the work we do, and what it starts with is a plain conversation about your numbers — including "not yet" when that's the honest answer. New to the topic? Start with what marketing automation actually is.