Standing service enquiry-sorting-and-reply-drafts-tested-monthly · revised 12 October 2026
Standing service
Keep your enquiry sorter and reply drafts tested every month
Each month 30 of your enquiries are labelled by you and compared with the sorter, the traps are re-run and 20 drafts are traced to your answer sheet. The send log must stay empty.
This starts a conversation by email. Nothing is charged, and nothing is run on your data, until we have agreed scope, access and terms with you in writing.
The responsibility you hand over
An enquiry sorter drifts as the business changes. The answer sheet goes out of date, new kinds of enquiry arrive, and a model update can shift behaviour silently. The one-off test passed on the day, and nothing re-runs it.
Who it’s for: Owner of a service firm or small sales team using the enquiry sorter and reply drafts.
Usually starts when: Your services, prices or wording have changed, and nobody has re-checked the sorter.
The result: Every month a sample of 30 of your enquiries is labelled by your person and compared with the sorter, the trap messages and instruction-bearing messages are re-run, 20 drafts are traced to your answer sheet and the send log is checked as empty.
What stays true, and what we do about it
No response-time guarantee is published for this new service. A target for delivering each list or record is agreed in writing before it starts, set to what a service at this stage can actually keep.
Hours are agreed in writing before the service starts. At launch the service is not staffed round the clock and runs on the cycle named here.
What must remain true
- Label agreement on the monthly sample stays at or above the pass mark of 26 of 30
- Every trap message is still marked "needs a person" and the send log is empty
What we watch
- A random monthly sample of enquiries with your labels
- The trap and instruction-bearing sets
- The answer sheet
When something happens
Scroll the table sideways to read it all.
| When | What we do |
|---|---|
| A month closes. | We run the sample, the traps and the trace and send the results. Linked job: Enquiries sorted with reply drafts |
| Label agreement falls below the pass mark. | We list the disagreements and propose a rule change for your approval. |
| A trap message is not marked "needs a person". | We flag it first and switch the draft step off in our recommendation. |
| An answer-sheet entry is past its review date. | We list it for its owner. |
We do on our own
- Run the monthly sample and traps
- Prepare the results
- Propose a rule change
We ask you first
- Any change to a rule, a threshold or the data covered
- Any request for wider access to a system
- Starting any fix, merge, send or write
We escalate to you when
- A trap message gets a draft
- The send log is not empty
- The answer sheet contradicts itself
How you know it held. Each month the results show per-label counts you can recompute from your labels, and the send log is empty.
How we keep it true
This service is never finished. Each period's record shows whether the thing it watches stayed true, and it continues until you end it.
Sort incoming enquiries and have a reply draft ready for you to send Job Optional
Each new enquiry from the form or mailbox you name is labelled, routed by your table and given a reply draft built only from your answer sheet. Nothing is sent.
Bought separately; not part of the monthly fee
Needed first, priced on its own page: the one-off sorter, drafts and tests.
What is included, and what is not
- The monthly results
- The trace results
- The list of stale answer-sheet entries
- Proposed label or routing changes
Included
- The sorter from the one-off job, for one form or mailbox
- A monthly label-agreement check on 30 randomly chosen enquiries
- The trap and instruction-bearing sets re-run
- A trace of 20 drafts to answer-sheet lines
- A list of answer-sheet entries past their review date
Not included
- Sending any message
- Editing the answer sheet: you own it
- Cold or promotional messages
- Regulated advice
- Real-time cover
How we know it’s done
Agreed with you before work starts. Each check produces evidence you keep.
On 30 randomly chosen enquiries labelled by your person, at least 26 get your label (a tripwire: it bounds nothing about future enquiries). Any severe error fails the month: a draft that states a price, date or policy not on the sheet, a trap message answered, or any message sent.
Evidence: The sample with your labels.
All traps are marked "needs a person" and the send log is empty.
Evidence: The trap results and the send log.
Every sentence in 20 sampled drafts traces to a line in the answer sheet.
Evidence: The trace table.
Sign-off. You read each monthly test report and tell us whether the rules, thresholds and test set still suit you. A period counts as delivered when you have received the record.
If it fails. If we cannot complete a period because an input is missing or has changed shape, we say so, explain what we need and do not guess. If the service is not working for you, you can end it at the end of any month.
When it fits, and when we stop
It fits when
- The one-off job has been run, or is bought with this service
- A person on your side labels 30 enquiries a month in about 30 minutes
- Real records are read only after a written agreement that covers how your records are processed and a separate access step you control; until then we test on invented or redacted data.
We stop and tell you if
- The answer sheet has no owner
- You want automatic sending
- No one labels the monthly sample
What could go wrong
The service only tests and reports. You switch the sorter off in your mailbox at any time.
Scroll the table sideways to read it all.
| Risk | How we handle it |
|---|---|
| A small monthly sample hides a drop | A sample of 30 is a tripwire. A severe error is reported in that month's report, and we recommend the draft step stays off until 150 in a row are clean. |
| The answer sheet is out of date | Stale entries are listed for their owner each month. |
| A model update shifts behaviour | The traps and the sample are re-run monthly and after any change you tell us about. |
A separate automated reviewer checks each monthly test report before it is sent, and no person on our side reviews unless your agreement names one. Your person labels the sample and approves every change. At launch the work is largely automated, and we say so.
Stays with a person
- Your person labels the sample and sends every reply
- You approve any change to rules or the answer sheet
Access we would need
- The monthly sample, sent by an agreed route
- A least-privilege mailbox app, revocable by you
Questions
Does this check accuracy for good?
No. It is a monthly tripwire on 30 enquiries. It does not promise a rate for all future enquiries.
Do you edit my answer sheet?
No. We list stale entries for its owner.
Do I need the one-off job?
Yes, it builds the sorter. It is priced on its own page and is not in the monthly fee.
Send an enquiry
Send us
- Changes to your services, prices or policies last month
- Any enquiry that was misrouted or answered wrongly
- Who labels the sample
Later, once you agree
- 30 randomly chosen enquiries each month, with your labels
- The current answer sheet
- A secure handoff route agreed before any record is shared
You keep the mailbox and the answer sheet. We read the monthly sample you send and write to Drafts only. Which service reads your records is agreed in writing before any real record is shared.
Email fallback: open your mail app
If website submission is unavailable, review and send the fallback email yourself. An email fallback is not a website receipt. Or write to hello@syntheticindustry.ai with “enquiry-sorting-and-reply-drafts-tested-monthly” as the subject.