SME proposal · Lynn Beauty
Turn an enquiry into a usable request.
Lynn Beauty · a proposed one-outlet receptionist pilot, with booking and multi-outlet expansion gates. Public services are real; workload, eligibility and all prices below are planning assumptions.
The operating loop
- 01CustomerApproved answer and service/time request
- 02ReceptionistPersistent fields and explicit status
- 03StaffAcknowledged request and confirmation
- 04OwnerAttendance, rework and net effort report
Start with answers, persistent requests and acknowledged human handover.
The problem to establish
Lynn Beauty publicly offers beauty services, booking and WhatsApp contact. Confirm the exact outlets, legal applicant, current calendar and operator coverage. The hypothesis is repeated enquiries interrupt staff and requests are reconstructed or missed—not that missed sales have already been proven.
Source: Lynn Beauty ↗
| Ask for | What it resolves |
|---|---|
| Two representative weeks of enquiry and booking records | Volume, repeated questions and actual outcomes. |
| Active handling and correction time | Whether the service can be economically useful. |
| Existing calendar and staff allocation | Whether booking integration is necessary or already solved. |
| Actual avoided overtime or unserved appointment demand | Whether capacity can become cash/contribution. |
Select the smallest complete workflow
One outlet, one number, approved answers and structured preferred-time/service intake. Staff confirm the next step. No autonomous appointment commitment, prices outside the approved list, refunds or treatment advice.
Selected estate: receptionist + one CRM/inbox + Jarvis exception brief + Academy SOP + GRC change/processing register. Existing website remains; AEO/content are optional separate growth work.
Add calendar availability, duration/resource rules, conflict protection, confirmation, cancellation and reminders only after the lean handover works. Booking is confirmed only by the authoritative calendar/approved person.
Require tests for simultaneous requests, stale slots, timezone, service duration, deposits and cancellations. Price this connector and its ongoing support separately; choose Cal.com or the existing booking platform after discovery.
Add outlet routing, location-specific knowledge, role access and cross-outlet reporting after the first outlet’s benefit is measured. Each legal applicant and deployment scope needs its own review; website references do not establish a single applicant entity.
Optional Synaptic/AEO creates reviewed service content. Add voice only with independently tested call transfer and costs. Neither expansion is automatically covered by the first chatbot grant.
The accountable system
| Party | Responsibility | Evidence |
|---|---|---|
| Lynn owner | Approves answers, outlet facts, prices and commercial boundary | Signed versioned facts; update owner. |
| Lynn operator | Acknowledges, confirms or rejects requests within staffed SLA | Receipt, acknowledgement and booking ID. |
| CyberG7 | Configuration, guarded replies, context, routing, monitoring and rollback | Critical tests, delivery/error log and export. |
| Joint review | Tracks attendance, handling/rework and productive use | Weekly cohort and renewal decision. |
Reprice the first step
The earlier S$12,000 subscription model failed the assumed capacity-only test. This iteration uses a bounded planning package; it is not a published CyberG7 price. Support scope and actual unit costs must be quoted.
| Planning line | First year | Treatment |
|---|---|---|
| Software access, 12 × S$350 | S$4,200 | Candidate basis only if exact approved package exists. |
| Configuration and role training | S$1,200 | Excluded from conservative grant model. |
| Messaging/model overage allowance | S$720 | Planning allowance; reconcile actual fees. |
| Full first-year cost | S$6,120 | Before GST and other unpriced costs. |
| Recurring annual cost | S$4,920 | Support assumed within the bounded subscription. |
| Conditional 50% × S$4,200 | S$2,100 | Unverified approval; default grant is zero. |
Illustrative capacity: 600 enquiries/month × three net minutes = 30 hours/month. At S$25/hour that is S$9,000 annual capacity, not a payroll saving. A separate realisation scenario assumes six previously paid overtime hours/month avoided (S$1,800/year), plus ten genuinely additional attended appointments/month at S$40 contribution (S$4,800/year). Total S$6,600 annual realised benefit; validate both and include new operator effort.
Planning assumptions in S$. Edit them to test the business case. Benefit means avoided expenditure plus incremental contribution, after new client-side supervision/rework costs. Capacity is reported separately.
How the calculation works
Cost = one-off cost + annual cost × horizon. Benefit = annual benefit × horizon. Grant = selected rate × assessed basis, limited by the activity ceiling where stated and never more than project cost. Horizon surplus = benefit − cost + conditional grant.
Steady-state payback = one-off cost ÷ monthly recurring surplus, without the grant. This compares operating value; it is not a cash-flow schedule for annual prepayment. The SME must finance the full first-year cost until any reimbursement. Payback excludes application/delivery delays, tax and financing. GST and any unpriced costs must be added to the real budget. No approved cost basis or award is established by this calculator.
Choose funding before commencement
Own-package chatbot funding requires exact IMDA pre-approval, which has not been verified for CyberG7. Use direct deployment with tax review, or a genuine verified approved principal. Do not commence this funded scope or pay a deposit before submission. The future product needs dashboard/export/security and same-solution reference evidence.
Source: GenAI customer engagement ↗ · IMDA vendor criteria ↗ · IMDA chatbot checklist ↗ · IRAS EIS ↗
Pilot and acceptance
| Stage | Work and pass condition |
|---|---|
| Before start | Applicant/funding decision; actual quote; two-week baseline; signed facts/operator. |
| Week 1 after authorised start | Persistent request intake, operator receipt, production tenant settings and reporting. |
| Week 2 | 100 agreed factual/critical cases; all agreed critical routes; takeover/rollback/export. |
| Weeks 3–6 | Four-week pilot: ≥90% complete intake, ≥95% acknowledgement within staffed SLA; ≥20% net labour reduction target. |
| End of pilot | Owner signs outcomes, cost and realisation. Add booking only if the next bottleneck and calendar integrity justify it. |
The customer sees a clear request status until confirmed. Never count an invitation, delivered alert or AI-assisted appointment as an incremental attended booking. Case-study permission is separate from purchasing the pilot.