Deciding who builds your automations, sales quoting, lead follow up, reporting, support tickets, whatever's eating your week, is a different call than deciding whether you need an automation agency at all. If the problem is one repetitive task, say a weekly report someone copies from three tools into a slide deck, a competent person inside the business can often wire that up in an afternoon with tools you already pay for, and no agency is needed. Past that point, the fastest way to sort the field is to ask what happens to the work if you stop paying them. Firms that resell a platform will lose the automation, because it lives inside someone else's software and stops the day the subscription does. Firms that build you a system will hand you something that keeps running, because you own it: the code, the credentials, the logic. Everything else in the vetting process follows from that one question. This article gives you the rest of the questions, in the order to ask them, so you can tell the difference on a first call rather than three months into a contract.
Two kinds of automation agency answer this search, only one builds you anything
Search for an automation partner and you'll get two very different businesses mixed into the same results page. The first is a platform reseller: they set up a tool such as a CRM's built-in workflow engine or a helpdesk's rule builder, configure your triggers, and bill you monthly per seat or per user. The second is a firm that builds automations wired directly into the tools you already run, your CRM, your inbox, your spreadsheets, that you own outright once the work is delivered.
Both call themselves an "automation partner." Both will show you a demo where a form fills itself in or a question gets answered instantly. The difference is what's underneath the demo. One is a configuration of software you're renting seats on, something in the range of EUR 1,800 a year for a mid-size team. The other is logic built for your specific process that has no per-seat license baked into it at all. This article is a vetting guide for the second kind.
The pricing model tells you which one you're talking to
Pricing is the cleanest signal you'll get before you ever sign anything with an automation agency. If a proposal charges per seat, per agent, or per user, you're paying for access to software the firm has resold you, and the price rises every time you add someone to the team. If a proposal quotes a fixed fee for a specific piece of automation, agreed before work starts, you're paying to own something once.
Read a pricing page the same way. Reseller pricing lists tiers: "Starter, 5 seats", "Growth, 15 seats." Builder pricing lists scope: which process gets automated, which tools it touches, what the handoff looks like. For what a fixed-price build actually looks like when it's written down, see our pricing. Ask directly: "If we hire three more people next year, does our monthly cost change?" A reseller's answer is yes. A builder's answer is no, because the automation doesn't care how many humans are logged into it.
Why you can't hire your way out of repetitive work
Most founders start this search after a bad hiring cycle. You hire someone to handle quotes, or leads, or reporting, they burn out on the repetitive half of the job within a couple of months, and you're back to interviewing. The hire isn't the problem. The repetitive half of most jobs doesn't need judgment, it needs a workflow: pull the data, apply the rule, fill in the template, log it, move on.
Here's a worked example, not a measured result: if a task takes 20 hours a month today and volume doubles, headcount needs to roughly double too, unless the repetitive part is handled by something other than a person. That's the actual argument for automating a process rather than hiring another person for it. Volume grows with revenue. Headcount doesn't have to.
This is also why the same search keeps coming up. A founder patches a repetitive process with a hire, the hire quits or plateaus, the backlog grows, and the founder starts looking for a firm to build the automation instead of a fourth attempt at hiring.
Six questions to ask on the first call
Ask these in order. The answers will sort resellers from builders faster than any case study.
- "Who owns the workflow logic if we stop working together?" A builder answer names you, in writing, and explains the handoff. A reseller answer talks about "your account" on their platform of choice, which isn't the same as ownership.
- "Can you show me a workflow with no per-seat license underneath it?" If every demo they show runs inside one platform's rule engine, they are a configuration shop, not a build shop.
- "What's the first thing you'd automate for us, and why that one?" A builder should be able to name a specific process by likely volume, for example: "quote requests, because a team your size is probably logging somewhere around 30 to 50 a week, and if each one takes 15 minutes to price by hand, that's a number worth checking, not one we'd claim to already know about your business."
- "What happens to turnaround time if our volume doubles next quarter?" You want to hear that the workflow scales without a proportional cost increase, not "you'd add another seat."
- "Do you write logic against our own tools, or only configure an existing app's automation rules?" Configuration inside a platform's rule engine is not the same as a workflow built against your CRM's API and your own data.
- "Can we talk to a client who took this in-house or moved it to another developer?" Builders don't mind this question because portability was the point. Resellers usually can't answer it, because there's nothing to move, it's a subscription.
If you want a partner and not a vendor, these six questions are the fastest way to find one instead of signing up for another software bill.
Red flags that mean you're being sold a reseller relationship
A few patterns show up consistently when an automation agency is reselling a platform rather than building you anything:
- They quote per seat, per agent, or per user rather than a fixed project fee.
- They can't clearly explain what happens to the automation the day you stop paying them.
- They steer every conversation toward one platform's app marketplace, because that's where their commission or partnership sits.
- Every case study they show operates inside the same one platform's ecosystem, with no example of a workflow that runs independent of it.
- The pitch uses "seamless" and "all-in-one" more than it names an actual piece of manual work they removed.
None of these are dealbreakers on their own; plenty of businesses are genuinely well served by a configured platform. The problem is being sold that as if it eliminates manual work, when what it actually does is move the manual work into a subscription you can't take with you.
What a built automation actually replaces
It helps to picture the mechanics instead of the pitch. Take order status questions as one example: a customer asks where their order is, someone opens a store admin, cross-references a carrier tracking page, and pastes a status back into the reply. A built automation does that lookup itself, pulling live tracking data and replying without a person touching it.
The same pattern shows up wherever manual work lives. A quote calculation that today means someone pulling prices from three spreadsheets and typing up a document becomes a workflow that assembles the same quote from the same data in minutes instead of an afternoon. A weekly report that means someone copying numbers out of five tools into a slide deck becomes a workflow that pulls the numbers and builds the deck itself. None of this lives as a macro inside a helpdesk or a formula in a spreadsheet someone forgot to update. It's wired directly into the tools that hold the data, which is what makes it a system rather than a faster way to do the same manual step. If you want to see the range of what this covers before you get anyone on a call, our use cases page runs through twelve of them.
What ownership needs to say in the contract, not just the pitch
"You own it" is a sentence anyone can say on a sales call. It only means something if it's written down, and if it's built somewhere you already control. Before you sign with any firm, get three specific things confirmed in the contract, not the pitch:
- Who holds the code and credentials. The workflow should live in your own n8n instance, your own Airtable base, your own Google Workspace, your own CRM, not a login the agency controls on your behalf. You should be able to open it yourself and see the logic.
- Whether you can take it to another developer. If the answer requires the original firm's continued involvement to function, that's a license with extra steps, not ownership.
- What infrastructure it runs on. A workflow built inside accounts you pay for and control is portable. A workflow built entirely inside an agency's own proprietary platform is not, no matter what the sales page calls it.
This is the actual test for owned by you: could you fire the firm tomorrow and still have a working system, because it's sitting in your own n8n account with your own credentials attached. If the contract doesn't answer that clearly, assume the answer is no.
A five-minute scoring sheet for comparing firms
Use this on your next three calls. Score each firm 0 to 2 on each line, 2 meaning a clear builder answer, 0 meaning a clear reseller answer.
- Pricing model: fixed build fee (2) versus per-seat or per-agent (0)
- Ownership answer: named, in writing, code and credentials in your own accounts (2) versus vague or platform-dependent (0)
- Demo: a real workflow wired to your own tools and data (2) versus a platform skin with rule-builder macros (0)
- Portability: can name a client who moved the system elsewhere (2) versus no example (0)
- Specificity: named your likely highest-volume manual task unprompted (2) versus generic "we automate your business" (0)
Anything scoring 7 or above out of 10 is worth a second call. Anything under 4 is a reseller, whatever they call themselves on the homepage.
Common questions
Is a built automation always more expensive than a subscription tool upfront?
Usually, yes, upfront. Ongoing, often no, because you stop paying per seat and stop paying someone to do work a workflow could close without them.
Can a firm build automation on top of the tools we already use instead of replacing them?
Yes. Many builds keep the existing CRM, inbox or spreadsheet as the system of record and add the automation around it. The ownership question is the same either way: ask who owns the logic and where it lives.
How long does it take to build an automation like this?
Most builds run a few weeks for the first workflow, not months, because they start with the highest volume manual task, not a full rebuild of how the business runs.
What if we already pay for a tool and don't want to switch?
You don't have to. An audit call, a two week Revenue Roadmap is how we run ours, maps what's manual inside your current setup and what a workflow would take off your team's plate without moving you off anything.
Start with the Revenue Roadmap
Two weeks, EUR 1,997, money back if it does not find at least 20 hours a month worth automating. The fee comes off your first build.
