> ## Documentation Index
> Fetch the complete documentation index at: https://docs.routiq.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Analytics & Reporting

> What's working, what isn't, and what it's worth. Conversion-by-channel, recovered-revenue tracking, cohort analysis, AI-conversation health checks.

<Info>
  **Why this matters.** Most AI receptionist tools show you "messages
  sent" and "cost." Routiq's analytics show you what those messages
  *did* — bookings made, revenue recovered, escalations handled,
  conversations Robyn struggled with. The distinction between volume
  metrics and outcome metrics is the difference between a tool you
  can defend to a board and one you can't.
</Info>

## What you can see

<CardGroup cols={2}>
  <Card title="Outcomes (the metrics that matter)" icon="bullseye">
    Bookings made, reschedules saved, cancellations recovered (via waitlist), recall conversion %, revenue recovered.
  </Card>

  <Card title="Volume (the metrics for ops)" icon="chart-bar">
    Calls handled, SMS sent, emails replied, widget conversations, average response time.
  </Card>

  <Card title="Channel comparison" icon="arrow-up-arrow-down">
    Side-by-side: which channel converts best for your clinic, by appointment type, by patient cohort.
  </Card>

  <Card title="AI conversation health" icon="heart-pulse">
    Where Robyn struggled — knowledge gaps, low-confidence handoffs, escalation rate by category.
  </Card>

  <Card title="Cohort analysis" icon="users-rectangle">
    Performance broken down by treatment type, time-since-last-visit, practitioner, age, location.
  </Card>

  <Card title="Cost vs. revenue" icon="scale-balanced">
    Outbound messaging cost vs. revenue from bookings recovered through that messaging — net contribution per campaign.
  </Card>
</CardGroup>

## Outcome metrics (the ones to watch)

| Metric                  | What it tells you                                                                                                                         |
| ----------------------- | ----------------------------------------------------------------------------------------------------------------------------------------- |
| **Bookings made**       | Total appointments Robyn booked across all channels                                                                                       |
| **Reschedules**         | Bookings saved that would have been cancellations                                                                                         |
| **Recall conversion %** | Of patients in a recall sequence, what fraction rebooked within 30 days                                                                   |
| **Revenue recovered**   | Estimated revenue from bookings that wouldn't have happened without Routiq (calculated against your configured per-appointment-type fees) |
| **Avg response time**   | How long from patient enquiry to Robyn first reply (target: \< 30s on call/SMS)                                                           |
| **Escalation rate**     | % of conversations Robyn handed to a human (target: 5–15% — too low = Robyn over-confident, too high = knowledge base too thin)           |
| **No-show rate change** | Pre-Routiq vs post-Routiq, calculated rolling 90 days                                                                                     |

## AI conversation health check

Distinct from outcome metrics — this is for spotting where Robyn needs help:

<CardGroup cols={2}>
  <Card title="Knowledge gaps">
    Top 10 questions Robyn couldn't answer this week, sorted by frequency. Click to add an answer to the knowledge base.
  </Card>

  <Card title="Low-confidence escalations">
    Conversations where Robyn escalated due to confidence threshold. Often indicates a fixable knowledge or instruction gap.
  </Card>

  <Card title="Repeated misunderstandings">
    Same patient asking the same thing twice = Robyn missed it the first time. Surfaced for review.
  </Card>

  <Card title="Drop-off points">
    Where in a booking flow patients hang up / stop replying. Helps spot UX or copy issues.
  </Card>
</CardGroup>

## What your dashboard looks like

> **Dashboard → Overview** (last 30 days)
>
> * **341** patient conversations handled
> * **97** appointments booked (28% conversion)
> * **24** reschedules captured (would have been cancellations)
> * **AUD \$14,600** revenue recovered (net of messaging cost)
> * **8.2%** escalation rate (within healthy band)
> * **3** unanswered questions added to knowledge gaps this week

## Setup

<Steps>
  <Step title="Configure per-appointment-type fees">
    **Settings** → **Account Setup** → **Appointment Types** → for each type, set average fee. Used to calculate "revenue recovered."

    Routiq pulls fee data from your PMS where available — only fill this in if your PMS doesn't have fees configured.
  </Step>

  <Step title="Set the no-show baseline">
    **Settings** → **Account Setup** → **No-show Baseline**. Enter your pre-Routiq no-show rate (or let Routiq calculate it from the first month of historical data). Used for the no-show-rate-change metric.
  </Step>

  <Step title="Choose dashboard time window">
    Default is rolling 30 days. Change to last 7 / 90 / 365 days, or a custom range.
  </Step>

  <Step title="Set up scheduled reports">
    **Settings** → **Reports** → **Scheduled Reports**. Email yourself or your team a weekly or monthly digest with the top 5 metrics.
  </Step>

  <Step title="Set alerts for outliers">
    **Settings** → **Alerts**. Get pinged when:

    * Escalation rate spikes > 25%
    * Knowledge gaps unanswered for > 14 days
    * Daily conversation volume drops > 50% (signals a forwarding outage)
    * Outbound cost exceeds budget for the month
  </Step>
</Steps>

## Exporting

* **CSV export** — any view, current filter applied
* **PDF report** — formatted for stakeholders, includes brand
* **Scheduled email** — weekly / monthly digest in inbox
* **API** — query metrics programmatically (contact us for API key)

## Common questions

<AccordionGroup>
  <Accordion title="How is 'revenue recovered' calculated?">
    For each booking Robyn made, the per-appointment-type fee × probability that the booking wouldn't have happened without Routiq (based on the patient's history and segment). Conservative by default — actual revenue is usually higher.
  </Accordion>

  <Accordion title="Why does my escalation rate matter?">
    Too low (\< 3%) means Robyn is answering things she shouldn't — a clinical-question slip-through risk. Too high (> 30%) means the knowledge base is too thin or the confidence threshold is set too cautiously. The healthy band is typically 5–15%.
  </Accordion>

  <Accordion title="Can I see analytics by practitioner?">
    Yes — **Dashboard** → **By Practitioner**. Bookings, recall conversion, escalation rate, and revenue recovered, broken down per practitioner.
  </Accordion>

  <Accordion title="How accurate is the no-show-rate-change metric?">
    Compared to your configured baseline. Most accurate after 90 days of Routiq running, when the baseline is well-anchored. Uses rolling 90-day windows to smooth seasonality.
  </Accordion>

  <Accordion title="Where can I see channel-by-channel breakdown?">
    **Dashboard** → **Channels** — every metric stratified by call / SMS / email / WhatsApp / widget.
  </Accordion>
</AccordionGroup>

## Next

<CardGroup cols={2}>
  <Card title="Patient segmentation" icon="users" href="/features/patient-segmentation">
    The cohorts feed analytics; analytics tells you which cohorts to invest in.
  </Card>

  <Card title="Lapsed-patient reactivation" icon="rotate-right" href="/outbound/lapsed-patient-reactivation">
    Where most "revenue recovered" comes from.
  </Card>
</CardGroup>