# Always-On Front Desk — Pricing and Margin

Last updated: build day. All costs are measured or quoted, not estimated, except where marked.

---

## 1. The price

| Tier | Setup (one time) | Monthly | Scope |
|---|---|---|---|
| **Single practice** | $2,500 | **$890** | 1 location, up to 4 providers, unlimited calls + texts |
| **Multi-location / DSO** | $1,800 per location | **$690 per location** | 2-15 locations, cross-location routing, group reporting |
| Add-on: Spanish language line | $400 | +$150 | Second language, same triage rules |
| Add-on: Outbound recall campaigns | $600 | +$240 | AI calls the 6-month recall list |

Month to month. No contract. 30 days notice to cancel. No per-minute or per-booking billing.

**Why $890 and not $1,500.** $890 is under four figures, which keeps it a manager-level decision
instead of a budget conversation. It is also unambiguously less than a fifth of a loaded salary,
which makes the comparison instant. There is room to raise it to $1,200 once there are three
reference customers with real answer-rate data.

**Why the setup fee is not zero.** Setup is where the actual work is: PMS integration, fee schedule
loading, provider templates, triage rule review. A free setup attracts practices that will not do
the two-hour session, and the two-hour session is the single biggest predictor of whether the
product works. The fee is a commitment filter as much as revenue.

---

## 2. What it replaces, in dollars

One full-time dental front office coordinator, US metro, fully loaded:

| Line item | Monthly | Basis |
|---|---|---|
| Base salary | $3,300 | ~$19-21/hr full time, US metro dental front office |
| Payroll tax + benefits | $900 | ~27% of base: FICA, unemployment, workers comp, health contribution |
| PTO, sick days, coverage | $300 | ~15 days/yr covered by temp or overtime |
| Turnover + retraining, amortized | $200 | Dental front desk turnover is high; ~$3,600 per replacement cycle amortized over 18 months |
| **Fully loaded** | **$4,700** | |
| **Hours actually covered** | **~40 of 168** | One seat, business hours only |

**Always-On Front Desk:** $890/mo for 168 of 168 hours.

| | Receptionist | Always-On Front Desk | Delta |
|---|---|---|---|
| Monthly | $4,700 | $890 | **-$3,810** |
| Annual | $56,400 | $10,680 | **-$45,720** |
| Cost per covered hour | $27.35 | $5.30 | **-81%** |
| Simultaneous calls | 1 | unlimited | |
| Coverage gaps | lunch, evenings, weekends, PTO, turnover | none | |

**Year one, including setup:** $10,680 + $2,500 = **$13,180** vs **$56,400**. Net **$43,220** saved.

**The honest caveat I state on every call:** most practices do not fire anyone. They keep the
person and stop paying them to be a switchboard. In that case the correct framing is not $3,810 of
savings, it is capacity: you just added 128 hours a week of phone coverage and your existing
front desk stopped doing the work they hate. The savings case applies to the practice that is
currently trying to hire a *second* seat, and that is the sharper prospect anyway.

---

## 3. The revenue side

Labor savings is the floor. Recovered production is the real number.

Reference practice: 120 inbound calls/month.

```
120 calls/month
  × 25%  unanswered (lunch, after hours, both lines busy)   =  30 missed calls
  × 20%  of missed calls were new patients                  =   6 lost new patients
  × $950 average first-year production per new patient      =  $5,700 / month
                                                            =  $68,400 / year
```

Recovery scenarios against the $890/mo cost:

| Recovery rate | Production recovered / mo | Net of $890 cost | Return on software spend |
|---|---|---|---|
| 25% | $1,425 | +$535 | 1.6x |
| 50% | $2,850 | +$1,960 | 3.2x |
| 75% | $4,275 | +$3,385 | 4.8x |

**Setup fee payback:** at 50% recovery, $2,850/mo of recovered production clears the $2,500 setup
in roughly **26 days**.

**Where this stops being true.** Below about 60 calls/month the recovered-revenue case is thin and
the pitch is purely labor cost. Below ~$500K production, decline the deal. Medicaid-primary
practices have high call volume and low per-patient value, so the multiplier collapses — also a
decline.

---

## 4. Cost to deliver (COGS)

Per practice, per month. This is the part that determines whether the business is real.

| Cost | Monthly | Notes |
|---|---|---|
| Telephony (SIP trunk + numbers + minutes) | $45 | ~1,200 min/mo at carrier rates |
| Speech to text | $28 | ~1,200 min at streaming ASR rates |
| Text to speech | $22 | ~35K characters/mo |
| LLM inference | **$0** | Runs on owned hardware. See below. |
| SMS (confirmations, intake links, reminders) | $18 | ~600 segments/mo |
| PMS integration / middleware | $35 | Per-practice connector |
| Monitoring, logging, transcript storage | $12 | |
| Error budget / reprocessing headroom | $20 | |
| **Total COGS** | **$180** | |

**LLM inference is $0 marginal.** The fleet is owned hardware, already paid for and already
running:

| Node | Hardware | Serving |
|---|---|---|
| blackwell | 3x RTX PRO 6000 | vLLM, Qwen3-Coder-30B-A3B |
| rig-256gb | 256GB | laguna-262k |
| rig-96gb | 96GB | 17 models incl. hermes3:70b |
| rig-36gb / rig-28gb | 36GB / 28GB | qwen3-coder:30b, deepseek-r1:32b |

~293GB VRAM total. Measured latency on blackwell for a live receptionist turn: **~550ms end to
end**, verified in-browser during this build. Electricity is a fixed cost that does not scale with
customer count, which is the entire structural advantage over a competitor paying OpenAI per token.

---

## 5. Margin

**Per practice, steady state:**

| | Amount |
|---|---|
| Revenue | $890/mo |
| COGS | $180/mo |
| **Gross profit** | **$710/mo** |
| **Gross margin** | **79.8%** |

**Setup engagement:**

| | Amount |
|---|---|
| Setup fee | $2,500 |
| My time (~6 hrs: 2hr session + 4hr config/test) | — |
| Effective rate | ~$417/hr |

**Contribution at scale (single-practice tier):**

| Customers | MRR | COGS | Gross profit/mo | Gross profit/yr |
|---|---|---|---|---|
| 1 | $890 | $180 | $710 | $8,520 |
| 5 | $4,450 | $900 | $3,550 | $42,600 |
| 10 | $8,900 | $1,800 | $7,100 | $85,200 |
| 25 | $22,250 | $4,500 | $17,750 | $213,000 |
| 50 | $44,500 | $9,000 | $35,500 | $426,000 |

Plus setup fees: 50 customers acquired over a year is an additional $125,000 in one-time revenue.

**Where margin degrades:**
- Support load. Budget ~1.5 hrs/month/practice after month two. At 25+ customers this needs either
  a support tier or a part-time hire, which is the first real cost step.
- Telephony overage at very high call volume. Practices over ~400 calls/mo should move to a
  $1,290 tier. Build that trigger into the contract now rather than renegotiating later.
- Custom PMS work. Open Dental is straightforward. Denticon and Curve cost more integration time.
  Price a $1,000 setup premium for anything outside the core five.

**Break-even:** COGS is variable and small, so the business is profitable from customer one. The
real constraint is my time, not unit economics: ~6 hours per setup means roughly 20 setups a month
is the practical ceiling for a solo operator before onboarding has to be productized.

---

## 6. Discounting policy

| Situation | Allowed | Never |
|---|---|---|
| First 3 reference customers | 50% off setup ($1,250) in exchange for a written case study and a reference call | Discount the monthly |
| Annual prepay | 1 month free (effectively $9,790/yr) | More than 1 month |
| Multi-location | Already priced at $690/location | Stacking with other discounts |
| "Can you do $500/mo?" | No. Walk. | |

The monthly price is the anchor for every future customer. Discount the setup fee, never the
monthly. A practice that will not pay $890/mo to replace a $4,700/mo seat is not price-sensitive,
it is unconvinced, and the answer to that is evidence, not a lower number.