How to Analyze Cancellation Insights in Alpaca Assistant

Last updated: August 26, 2026

How to Analyze Cancellation Insights in Alpaca Assistant

Cancellation Insights is Alpaca’s organization-wide view of lost sessions — where they happen, who’s affected, and what they cost. This guide walks you through reading the dashboard’s KPIs, splitting family vs. staff cancellations, working the at-risk tables, reading revenue impact, drilling into individual cancellations, exporting data, and clearing overdue sessions.

Open the dashboard

  • In the left sidebar (under Organization), click Cancellation Insights.

  • The page header reads “Cancellation Insights” with your org name and the tagline “where sessions are lost, who’s affected, and what it costs.”

Cancellation Insights in the left sidebar

Set the time window first — it scopes everything

Every number, chart, and table on the page obeys the controls at the top. Set them before you read anything.

  • Preset windows: click 30d, 90d, or 180d (top-left).

  • Custom range: click the date field to the right of the presets (e.g., “05-22-2026 – 08-20-2026”) and choose your own start and end dates.

  • By session date vs. By cancellation date — the easy-to-confuse toggle:

    • By session date (default) buckets each cancellation by the day the session would have happened.

    • By cancellation date buckets by the day the cancellation was entered.

    • Use “by session date” for “what did we lose that week”; use “by cancellation date” for “what got cancelled recently.”

  • Include assessments — off by default; tick it to fold assessment appointments into the numbers.

  • Data as of… (right side) shows the last refresh. The dashboard is event-driven and refreshes roughly every 3 hours — click the circular refresh icon to pull the latest.

Time-window controls and the session-date / cancellation-date toggle

1. Cancellation rate and KPIs

Five cards sit directly under the controls. Each shows the metric for your window plus its change “vs. prior period” (an equal-length window immediately before).

  • Cancellation Rate — cancelled ÷ total scheduled sessions. The mini-sparkline shows the trend.

  • Family vs. Staff — two percentages, family / staff (see section 2).

  • Late-Cancel Rate (<24h) — share of sessions cancelled within 24 hours of start. (24h is the current fixed threshold; a configurable per-clinic SLA is planned.)

  • Hours Lost — total clinical hours in cancelled sessions.

  • Revenue Left on the Table — estimated dollars lost (see section 5).

  • Watch the arrows: an orange +9.8 pts / +245% is a rise (worse); a green −1.2 pts is a drop (better).

  • Read the caveat under the cards: sessions with no rate on file “couldn’t be priced,” so revenue is understated in those ranges — treat high-unpriced periods as coverage gaps, not real dips.

The five KPI cards and the time-window controls

The Weekly cancellations by reason chart below the cards plots each week’s cancellations as stacked bars with the cancellation-rate line overlaid (right axis). Hover any bar for a tooltip breaking out Family / Staff / Other / Unknown and that week’s rate.

Weekly cancellations by reason chart with hover tooltip

2. Family vs. staff cancellation split

  • The Family vs. Staff KPI card reads as family % / staff % (e.g., 4.8% / 5.1%).

  • In the chart, the stacked colors are the same split: Family (blue), Staff (orange), Other (dark gray), Unknown (light gray).

  • Important — it’s approximate: the split comes from the reason a staff member picks at cancellation time, not a verified actor. Read it as directional, not exact. (This caveat is printed right under the chart.)

3. Overdue sessions

Scroll to the bottom section. Overdue sessions are scheduled sessions with no cancellation and no completed note — probable no-shows. They are kept out of every metric above and never folded into the headline rate.

  • They appear only after a one-week buffer from the session date (leaving room for late or retroactive documentation).

  • Three summary cards: Overdue Sessions (count), Hours, and Estimated Revenue.

  • The table lists Session Date, Client, RBT, Hours, and Note Overdue (days past due).

Overdue Sessions summary cards and table

Bulk-clear the backlog:

  • Click Cancel [N] overdue sessions (top-right of the section). It works in batches of 50 — if any remain afterward, run it again.

  • A confirmation explains: these billable, never-started, undocumented sessions are cancelled on the clinicians’ behalf and recorded as staff cancellations; families are not notified; a cancellation can be reversed from the schedule if needed.

  • Click Keep them to back out, or Cancel [N] overdue sessions to proceed.

Cancel overdue sessions confirmation dialog

4. At-risk families and authorization utilization

Two ranked tables surface who’s driving cancellations. Only clients/RBTs with 10+ scheduled sessions are ranked.

  • RBTs at risk — ranked by cancellation rate (diagnostic, not a performance score). Columns: RBT, Total Sessions, Total Cancellations, By Staff, By Family, Late Cancels, Cancellation Rate, Revenue Impact.

RBTs at risk table
  • Families at risk — combines two independent signals: how often sessions are cancelled and how much of the authorization is used.

    • Auth. Utilization shows a percent and bar of authorized units consumed. Low utilization next to a high cancellation rate flags a client at risk of under-using — and losing — their authorization.

    • The remaining columns mirror the RBT table (Client, cancellations By Staff / By Family, Cancellation Rate, Revenue Impact).

  • Use Rows per page and the pager at the bottom of each table to move through everyone; the “X shown” label tells you the total.

Families at risk table with the Auth. Utilization column

5. The revenue-impact column

  • Appears on both at-risk tables. It is the estimated dollars lost to that person’s cancellations in the selected window.

  • Calculated from each cancelled session’s scheduled duration × the reimbursement rate on file for that date — so it is a predicted value.

  • A $0 usually means those sessions had no rate on file (non-billable or missing rate), not that they were free — pair it with the unpriced-sessions caveat at the top of the page.

Revenue Impact column on the RBTs at risk table

6. Drill-in and export

Drill into any number: underlined counts (Total Cancellations, By Staff, By Family, Late Cancels) are clickable.

  • Click one to open “Cancelled sessions — [name],” listing every session behind that count for the window: Session Date, Client, RBT, Reason (Family / Staff / Other), a Late pill (cancelled within 24h), Hours, and Revenue.

  • Close it with the × in the top-right corner.

Drill-in modal listing every cancelled session behind a count

Export to CSV: each section (RBTs at risk, Families at risk, Overdue) has its own Export CSV button in its top-right corner.

  • The export includes all rows in the current window — including rows hidden by paging, not just what’s on screen.

Export CSV button on the RBTs at risk table

Quick FAQs

Why is Revenue Impact $0 when there were cancellations?

Those sessions had no reimbursement rate on file for their date, so they couldn’t be priced. The dollars are understated, not truly zero — treat it as a coverage gap.

What counts as a “late” cancel?

Any cancellation within 24 hours of the session start. That threshold is currently fixed; a per-clinic configurable SLA is planned.

Do overdue sessions affect my cancellation rate?

No. Overdue (never-started, undocumented) sessions are tracked separately and excluded from every KPI and the headline rate until you act on them.