Quick Takeaways
- Repeatable internal workflows are the most undervalued asset most solo builders are sitting on right now
- The gap between “automation I built for myself” and “product I sell” is mostly packaging, not engineering
- The fastest path to first revenue is selling the workflow as a done-for-you service before wrapping it in software
- Platform costs (Zapier, Make, n8n, Airtable) are low enough at small scale that margin is real from day one
The Opportunity Most Builders Miss
Here is what tends to happen: a developer or operator builds a workflow to solve a personal bottleneck — maybe it’s an intake form that routes leads to a CRM, triggers a follow-up email sequence, and logs everything in Airtable. It works. They move on. Six months later they’re building the same thing for a client and charging a flat project fee, then watching the client struggle to maintain it alone.
The missed play is in the middle. That workflow — the one you built once and forgot about — is a product waiting to be extracted.
In 2026, the tooling makes this genuinely viable at small scale. Zapier connects over 9,000 apps with task-based pricing starting around $30 a month. Make and n8n offer comparable power at lower per-run costs. Airtable, despite its pricing friction at enterprise levels, works well as a lightweight backend for workflows serving small teams. The infrastructure cost of running a productized workflow for a handful of customers is often under $100 a month. That means your first few customers are almost pure margin once you’ve absorbed the build.
> The real asset isn’t the automation. It’s the operational knowledge baked into it — the edge cases you already solved, the error handling you already built, the logic nobody else knows to ask for.
This is the moat that most people advertising “AI automation agencies” on LinkedIn are skipping. They’re selling the tool, not the solved problem. Your advantage is the solved problem.
Three Shapes This Can Take
1. Done-for-you service. You run the workflow on behalf of the client. They send you inputs; you return outputs. This is the easiest starting point because the workflow lives on your infrastructure and you control quality. It’s not fully passive — you’re still in the loop — but it’s often the right first shape because it lets you charge premium prices while you figure out where the edge cases actually live.
2. Done-with-you setup + handoff. You build the workflow in the client’s Zapier or Make account, document it thoroughly, and train them to maintain it. One-time project fee, typically $500 to $3,000 depending on complexity. This works well for clients who want ownership. Your risk is they call you for support forever unless you charge a retainer.
3. Managed automation product. You run the infrastructure, clients pay monthly. This is the SaaS direction — lower per-customer revenue but compounding. A narrow product serving a specific use case (say, automated proposal generation for independent consultants, or intake-to-invoice routing for solo accountants) can reach $2,000 to $5,000 MRR with 20 to 40 customers before you’ve written a line of custom software. Most of the stack is glued together with existing tools.
The progression matters. Start with shape one, identify the parts clients ask about most, and let that tell you what to harden into shape three.
Who This Is For — And Who Should Skip It
Built for: Developers, ops leads, and technical freelancers who have already built automations for themselves or clients and have a repeatable pattern emerging. Also strong for people currently working in a specific industry who know the internal workflows cold — legal intake, insurance claims processing, real estate transaction coordination, accounts payable. The more domain-specific the workflow, the less competition and the higher the willingness to pay.
Also viable for: Consultants who are manually doing work that follows the same steps every engagement. If you’re producing the same deliverable repeatedly, you have a workflow that can be partially automated and then resold.
Skip it if: You haven’t actually built and run a workflow end-to-end yourself. People who theorize about automation and then sell setups they’ve never stress-tested create a specific kind of client damage — automations that work on the happy path but break on the first real exception. The research is consistent here: teams that automate without planning for how work actually happens in production produce systems that erode trust fast.
Also skip it if you’re hoping this is passive income from day one. It isn’t. The first three to six months are active: building, troubleshooting, documenting, supporting. Passive comes after you’ve systemized the support layer too.
A Concrete Starting Sequence
- Audit what you already run. List every recurring task you’ve automated in the last 12 months — for yourself or clients. Score each one: how often does it run, how much manual time does it save, and how specific is the use case?
- Pick the most transferable one. Not the most impressive technically — the most transferable. A workflow any small business in a specific niche could use is more valuable than a clever one-off.
- Offer it as a service to two people before you build anything new. Charge $200 to $500 for a setup and one month of support. Use real clients to find the real failure modes.
- Document obsessively during those first engagements. Every support question is a product gap. Every exception you handle manually is a feature.
- Price the managed version at $97 to $197 per month per client once the workflow is stable. At 15 clients, that’s $1,500 to $3,000 MRR on infrastructure you’re already running.
- Raise prices on new clients every five customers. Early customers are paying for your learning. Later customers are paying for a proven system.
The Real Tradeoffs
Platform dependency is the quiet risk nobody talks about enough. If your entire product runs inside Zapier and Zapier changes pricing or deprecates an integration, your margin collapses overnight. Diversifying across Make or n8n for critical paths reduces this — n8n in particular allows self-hosting, which gives you cost control at scale.
There’s also the human checkpoint problem. Automation earns trust slowly and loses it instantly. The fastest way to damage a client relationship is to automate something that carries financial, legal, or reputational weight — invoice amounts, customer-facing communications, compliance documents — and then call it hands-off. Build human review steps into anything that touches those categories, even if it feels like it undermines the automation story.
Finally: customer acquisition doesn’t solve itself. A workflow product with no distribution is a hobby. The operators who win here typically pick one narrow niche and become the obvious answer in one community — a subreddit, a Slack group, an industry association newsletter. Generic “I build automations” positioning is invisible. “I build client onboarding systems for independent financial planners” is a sentence someone can act on.
Bottom Line: The workflow you built and forgot about is probably your best product idea — extract it, validate it with two paying clients, and build the managed version only after you’ve run it in production long enough to know where it breaks.