Process
Start small. Get it right. Then grow.
Three stages, and a deliberate refusal to automate four things at once. Here is what each stage involves, what we need from you, and what you get at the end of it.
Why the first scope is deliberately narrow
A large automation built in one pass is close to impossible to debug. When something changes — your hours, a price, a tool's API — the failure shows up somewhere in the middle and there is no clean way to isolate it.
Our own dental reception build is sixteen separate workflows for this reason, not one big one. Intent routing, availability checks, request creation, confirmation messaging and escalation are each independently testable, which means each is independently fixable. The same principle applies to a first engagement: one workflow, working, that you can actually verify.
The three stages
Stage 1: Understand
We map how enquiries, calls and appointments actually move through your business today.
Usually one or two conversations
What we need from you
- Half an hour with whoever actually answers the phone.
- Your real opening hours, including the days they differ.
- An honest account of what currently goes wrong.
What you get
- A written description of the workflow as it runs today.
- A scope for exactly one workflow, with its boundaries stated.
- A plain answer on whether automating it is worth doing at all.
Stage 2: Automate
We design and configure automation around those flows, starting with a narrow, well-defined scope.
The bulk of the work
What we need from you
- Access to the tools involved, or an administrator who has it.
- Your escalation rules — what must always reach a person.
- Someone to review the test runs with us.
What you get
- The workflow built and tested outside your live environment first.
- An explicit failure path, so a workflow that stops tells someone.
- The handover points written down, not implied.
Stage 3: Improve
We review how the workflow performs in practice and refine it as your requirements change.
Ongoing, while it matters
What we need from you
- The cases where it got something wrong. These are the useful ones.
- A heads-up when your hours, prices or staffing change.
What you get
- Fixes for the edge cases real usage surfaces.
- Refinements to wording, rules and hand-offs.
- A recommendation on whether to widen the scope — sometimes that is 'not yet'.
What we will tell you that a vendor usually will not
Sometimes the answer is that the process is not worth automating — it happens twice a month, or the rules change too often to encode, or the real problem is that two people disagree about who owns it. We would rather say that in the first conversation than build something you stop using.
Every engagement follows the process above, starting narrow and expanding only once the first workflow is working. More about the team on the about page.
Stage one is a conversation, not a contract.
Tell us what currently goes wrong and we will tell you whether automating it is worth your time.
