The Four Key Metrics
Each KPI tile on your dashboard measures a specific dimension of your virtual receptionist's performance.
Total Calls
What it shows
The total number of guest calls your virtual receptionist has handled since going live. Every call — whether it lasted 10 seconds or 10 minutes — counts as one.
How it's collected
When a guest hangs up, our voice platform automatically sends the call details to our cloud backend, which saves a record to your database. Each saved record increments this count.
Why it matters
This is the baseline volume metric. The trend arrow shows whether call volume is growing or shrinking week over week.
Example
247 calls handled all time
Hours Saved
What it shows
The total amount of front desk staff time your virtual receptionist has absorbed. Every second it is on a call is a second your staff didn't have to be.
How it's collected
Our voice platform automatically reports the exact duration of each call in seconds once the conversation ends. This metric adds up every call's duration and converts the total into hours.
Why it matters
This is your ROI number. Front desk labor costs money — every hour your virtual receptionist handles frees your team for higher-value work.
Example
14.3h of front desk labor saved
Avg Duration
What it shows
The average length of a guest call, calculated across all calls that have a recorded duration. Shown as minutes and seconds.
How it's collected
Each call's duration is automatically recorded in seconds. This metric divides the total seconds across all calls by the number of calls to get the average.
Why it matters
Unusually long calls can signal that the virtual receptionist is struggling with certain request types — a cue to review those transcripts and improve its configuration.
Example
2m 18s average per call
Completion Rate
What it shows
The percentage of calls where your virtual receptionist successfully handled the guest's request from start to finish. A higher rate means more calls resolved without issues.
How it's collected
Every call is automatically assigned a status when it ends. A call is marked completed when the conversation reached a natural conclusion and the guest's request was handled. Calls that dropped or ended in error are tracked separately and not counted as completed.
Why it matters
This is your reliability metric. A completion rate above 90% means your virtual receptionist is consistently handling guest requests end to end.
Example
94% of calls fully resolved
How Data Gets to Your Dashboard
Every metric follows the same automatic pipeline — no manual input required.
Guest Calls
A guest calls the hotel's phone number through our phone network
Concierge Answers
Virtual guest concierge handles the full conversation
Call Analyzed
Our voice platform analyzes the call and extracts duration, transcript, issue category, and outcome
Data Transmitted
Call data is securely and automatically sent to our cloud backend
Data Stored
Our encrypted database saves a permanent record of every call
Dashboard Updates
Your dashboard refreshes automatically every 60 seconds
Reading the Guest Interaction Log
Each row in your call log is one guest conversation. Here's what every element means.
+1 (954) 555-0182
Maria Rodriguez
Completed
Reservation
⏱ 2m 14s
"Hi, I'd like to check availability for the weekend of..."
Today
2:41 PM
| Element | What it means | Where it comes from |
|---|---|---|
| Left color bar | Guest sentiment during the call — green = positive, amber = neutral, red = negative | Analyzed automatically by our sentiment analysis system after each call |
| Phone number | The caller's phone number, captured automatically when the call connects | Captured automatically from our phone network at the start of each call |
| Guest name | The guest's name if they stated it during the call and the virtual receptionist extracted it | Extracted automatically from the conversation by our transcript analysis system |
| Status pill | Whether the call completed successfully or encountered an issue | Reported automatically by our voice platform at the end of each call |
| Category pill | The type of request — Reservation, Check-in/Out, Maintenance, Tiki Bar, or General Inquiry | Extracted automatically from the conversation transcript after each call ends |
| Duration pill | How long the call lasted in minutes and seconds | Measured automatically for every call and recorded when the call ends |
| Transcript preview | The first 90 characters of the conversation. Click the row to read the full transcript with chat-style bubbles. | Stored securely word-for-word in our encrypted cloud database |
Frequently Asked Questions
Quick answers to common questions about how the dashboard works.
The dashboard automatically refreshes every 60 seconds in the background — you don't need to reload the page. New calls appear within one minute of the guest hanging up. The live indicator in the top-right corner pulses green to confirm the connection is active.
Issue categories are determined automatically by our transcript analysis system. After each call ends, the full conversation is analyzed and the guest's request is classified into one of five categories: Reservation, Check-in/Check-out, Maintenance, Tiki Bar, or General Inquiry. No manual tagging or review is required.
Each KPI tile has a small trend badge in the bottom-right corner. It compares the current 7-day window to the previous 7-day window. A green ↑ means the metric improved, a red ↓ means it declined, and → means it stayed flat. This helps you spot patterns week to week without having to do the math yourself.
Yes. All call records are stored in our encrypted cloud database, hosted on enterprise-grade infrastructure with encryption at rest and in transit. Only users with authorized access to this dashboard can view call records. No call audio is ever stored — only the text transcript and call metadata.
Yes. The Guest Interaction Log has an Export button that lets you download the full call log — including phone numbers, names, durations, categories, and transcripts — as either a CSV or Excel file. You can use the date range filter first to export only a specific time period. The busiest hours chart also has its own separate export.