DEV Community

Redman WorkflowAlpha
Redman WorkflowAlpha

Posted on

Every automation I sell has the same 7 parts. Part 6 is what clients pay for.

Every automation I sell has the same seven parts. Here is the one that matters most: the failure-handling section.

The problem with most automation

People buy automation to stop doing a manual task. Then the automation breaks at 2 a.m.
and they are doing the manual task again, only now they also have to debug the thing that
was supposed to save them.

I ran into this over and over with client work. The automation worked perfectly in the
demo and failed in the real world because nobody wrote down what happens when the input is
wrong, the API is down, or the client sends a message at midnight.

The fix: write the failure modes first

Every SOP I write now has these seven parts:

1. Purpose       -> one sentence, what this replaces
2. Triggers      -> what starts it
3. Inputs        -> exactly what it needs, and where it comes from
4. Steps         -> the happy path, in order
5. Decision      -> the branch point (if X, then Y)
6. Failure       -> what happens when each input is wrong
7. Outputs       -> what the client actually receives
Enter fullscreen mode Exit fullscreen mode

Parts 1-5 are what anyone can write. Part 6 is what clients pay for. It answers the
question they were afraid to ask: "what happens when this breaks?"

A real example: appointment reminders

A clinic in Tuzla wants automated SMS reminders. Simple, right?

The SOP I would write has a failure section that says:

FAILURE HANDLING
- Patient has no phone number -> skip SMS, add to "needs phone" list, notify front desk
- SMS provider returns 5xx -> retry once after 60s, then queue for manual send
- Patient replies "STOP" -> mark opted out, stop all future messages, log it
- Appointment cancelled -> cancel the reminder, do not send
- Reminder already sent -> do not send twice (idempotency check on appointment_id)
- Clinic closed on that day -> skip, do not reschedule automatically
Enter fullscreen mode Exit fullscreen mode

That is the difference between an automation that works and an automation that is safe to
run unattended.

What I sell

A pack of 12 of these SOPs, one for each automation a small business actually buys:

  • Lead capture into a CRM
  • Cold outreach with follow-up
  • Reply detection
  • Invoice reminders
  • Appointment reminders
  • Weekly content batch
  • Ticket triage
  • Status reports
  • Meeting actions
  • Competitor monitoring
  • Nightly backup
  • Client onboarding

Each one is Markdown (editable) + a styled A4 PDF + a commercial licence. Use them with
your own clients.

The point

If you write automation, write the failure handling. Not after it breaks — before you
ship it. That is the part that makes it sellable, and the part nobody enjoys writing.


Which of these 12 would your business actually use first? That is the one to start with.

Top comments (0)