Almost every automation problem I see can be solved inside GoHighLevel — CRM, funnels, SMS/email, Conversation AI, Stripe. It's fast to build on, it's a platform the client actually owns, and it covers the overwhelming majority of what a service business needs to run its busywork. But every so often, a request comes in that GoHighLevel genuinely isn't the right tool for, and knowing the difference matters.
The signal that says "stay in GoHighLevel"
If the job is coordinating existing data and existing communication channels — follow-ups, reminders, booking, retention, recovery — that's squarely GoHighLevel's job. It's built for exactly this, it's fast to iterate on, and there's no reason to reinvent it with custom code. Ninety percent of automation requests fall here, and building a bespoke app for this kind of work is usually over-engineering: more surface area to maintain, more cost, no real benefit over a well-built workflow.
The signal that says "build a real app"
The tell is when the ask is actually a product, not a workflow — a genuinely new interactive experience the end user interacts with directly, not just a sequence of messages triggered behind the scenes. An AI interviewer that walks someone through a structured conversation and produces a document. A calculator, a configurator, a tool with real UI state. GoHighLevel's funnel builder can fake some of this, but it starts fighting you the moment the interaction gets nonlinear or needs real application logic.
What "custom" actually means here
When that signal shows up, the right move is a standalone app — typically Next.js with the Vercel AI SDK and OpenAI — that can live on its own or sit alongside the GoHighLevel system, feeding data back in where it matters. It costs more to build than a workflow does, so it's worth being honest about when it's actually warranted versus when a client just assumes "custom" sounds more impressive than "automated."
The decision, simplified
Ask one question: is this coordinating something that already exists, or creating something new that a user interacts with directly? Coordination stays in GoHighLevel. Creation gets its own app. Getting this call right up front saves real money — building a custom app for a job GoHighLevel already does well is wasted spend, and forcing a real product into a workflow builder produces something that never quite works right.