Twilio voice reporting
Client call performance, made clear
Call volume, answer rate, outcomes, talk time and connectivity cost — plus the numbers, routes and readiness behind them, for a single client account. Sign in to open the voice report, or preview what it covers.
Twilio voice reporting
Voice is the channel most agency reports skip, because the numbers sit in a provider console nobody wants to hand a client. This dashboard organizes them the way an account team reviews every other channel: call volume and answer rate first, then outcomes, quality, talk time and connectivity cost, with the numbers, routes and readiness behind them — all scoped to one client.
The report starts where the client’s question starts: how many calls happened, how many were answered, and how many failed outright — day by day, for whatever window you set. A summary band carries the totals for the range and a daily table underneath shows their shape, so a quiet Tuesday, a spike after a campaign launch or a weekend with nobody picking up is visible instead of being averaged into a single monthly figure.
The date range is yours to choose, up to 366 days, and every day is read in the reporting timezone your workspace has configured rather than the server’s clock — so a day in the report is the day the phone actually rang. Answer rate for the whole range sits in the summary band beside calls, answered calls, failed calls, talk time and cost, so it stops being something a coordinator recalculates in a spreadsheet on the morning of the review.
A call total hides the story. The outcome breakdown splits the range into answered, unanswered and failed calls, each with the share of the total it accounts for, so a soft week can be diagnosed instead of guessed at. A rising share of failed calls is a different conversation — and a different fix — from a rising share of unanswered ones, and the breakdown separates the two before anyone has to speculate in front of the client.
An answered call is not automatically a good call, so a dedicated quality view carries the Voice Insights measurements for the same days: average MOS, jitter in milliseconds and packet loss. Those fields appear when the client’s account is entitled to Voice Insights and the measurements have been imported; when it is not, the view says so plainly and the call facts remain available. Nothing is invented to fill the gap.
Agencies running an AI booking agent on a client’s phone number end up with a channel that dials all day and reports almost nothing. Those are Twilio calls, so they land in the same daily facts as every other call on the account: volume, answered calls, the outcome mix and talk time, over any range you pick. That is the honest scoreboard — whether the agent is reaching people at all, and what happens when it does — rather than an anecdote about the two conversations somebody happened to overhear.
Caller identity is part of that result, which is why a trust view tracks the registrations behind it: the compliance profile, SHAKEN/STIR, Voice Integrity, CNAM and branded calling, each with its canonical state and when that state last moved. Whether a registration is approved, still in review or rejected is what decides how a call is labelled on the handset, so those states belong beside the answer rate rather than buried in an email thread. The conversation itself is authored in the booking-script library; this report covers what the calls did.
A client is rarely one phone number. The numbers and routes view lists every number the account has held — its masked value, whether it is the primary line, its number state, the route behind it (a direct webhook, a TwiML application or a SIP trunk), that route’s configuration state and when it last synced. Numbers stay masked in the interface and in the export, so the report can be opened in front of a client without putting raw contact data on the screen.
Above all of it sits the sync banner: the current reporting state, when the last successful snapshot published, whether what you are reading has gone stale, and whether an earlier day is still incomplete and being recovered. When a refresh fails, the last good snapshot remains readable and the banner explains the shortfall instead of presenting an empty report as a quiet month. You always know how much weight the number in front of you can carry.
The usage and cost view puts duration and connectivity cost on the same daily row as the call count, formatted the way people actually discuss them — hours and minutes rather than a column of raw seconds, and currency rather than micro-units. Talk time is the honest measure of how much conversation a channel produced: a call count on its own flatters a burst of hang-ups and undersells a fifteen-minute booking that closed.
Cost is reconciled against the provider’s own usage records for the same window rather than estimated from a rate card, and it publishes together with the call facts as one snapshot, so the two halves of a day can never disagree. With calls, minutes and cost on one row, cost per call is arithmetic you can do in front of a client, and a month where spend climbed faster than talk time shows itself immediately.
Twilio reporting is scoped to a single client identity: choose the client in the switcher and the report loads that client’s verified voice facts and nothing else. There is no blended workspace view to screenshot by accident, and no path for one account’s numbers to surface inside another account’s review. The client, the view and the date range are all written into the URL, so a link pasted into a channel reopens exactly the report you were looking at.
Six views carry the whole review — overview, calls, quality, usage and cost, numbers and routes, and trust — inside the same white-label workspace as the client’s projects, files and CRM, presented under your agency’s branding. Where your role permits it, the range on screen exports as a masked CSV for a deck or an invoice reconciliation. Connecting and provisioning the client’s Twilio account stays a separate setup step, owned by the people who should own it.
A client's voice activity from Twilio's own facts: daily call volume, answered calls and answer rate, call outcomes, quality, usage with reconciled connectivity cost, the numbers and routes involved, and trust state. It is organised as six views for one client at a time, and it is the one report here reading real production facts rather than a deterministic preview engine.
Answer rate is answered calls divided by total calls in the selected range, reported daily so a bad week shows up as a shape rather than disappearing into an average. The outcome breakdown then splits the same total into answered, unanswered and failed, so a falling answer rate can be traced to the outcome that actually moved.
Yes. Automated booking-agent calls are reported on the same footing as the ones a human team takes — volume, outcome and duration — which is the first thing a client asks for once an agent is answering their phone. Without it, the agent is the only part of the account nobody can put a number against.
The usage and cost view puts daily call counts, talk duration and reconciled connectivity cost in one table. Reconciled means the figure is published from the reporting ledger once the underlying facts settle, rather than estimated at call time, so the number in the report is one an agency can bill against or check against an invoice.
No. The marketing page you are reading is public; the call data behind it is not. Twilio reporting is scoped to one immutable client identity and served through a protected endpoint, so figures load only for a signed-in agency user with rights to that client, and none of it is projected into the client portal.
