What Is a Predictive Dialer? The Call Center Operator's Guide

If you run an outbound call center and you are not using a predictive dialer, you are paying agents to listen to ringing. That is not an exaggeration. In a manual or preview dialing environment, agents spend 60 to 70 percent of their time waiting — waiting for the phone to ring, waiting for someone to answer, waiting through voicemails, waiting through disconnected numbers. A predictive dialer cuts that idle time to under 5 percent by algorithmically managing the outbound call volume so that an agent is connected to a live human the moment they become available.

This guide is written by SIPNEX, an FCC-licensed carrier that provides SIP trunks for predictive dialers and VICIdial operations specifically. We run VICIdial on our own network. Every trunk we sell to a dialer operator has been tested on the same infrastructure we use for our own campaigns. What follows is not a software review — it is the carrier-side perspective on how predictive dialers work, what they demand from the underlying trunk, and why operators who blame their software for bad performance are usually looking at the wrong layer.

What a predictive dialer does

A predictive dialer is software that places outbound calls on behalf of a pool of agents, using statistical algorithms to determine how many calls to place simultaneously so that a live human is ready to connect to an agent the moment that agent finishes their current conversation. The dialer does not wait for an agent to become free before initiating the next call. It predicts when the agent will become free — hence the name — and begins dialing in advance so the connection is ready the instant the agent’s line opens.

The math behind predictive dialing is more sophisticated than most operators realize. The dialer maintains a real-time model of several variables: the current answer rate (what percentage of outbound calls are being picked up by a live person), the average call duration (how long a typical conversation lasts once connected), the average wrap-up time (how long the agent spends on after-call work before they are ready for the next call), and the number of agents currently available or about to become available.

From these inputs, the dialer calculates a dial ratio — the number of simultaneous outbound calls to place per available agent. If the answer rate is 20 percent (one in five calls gets picked up), the dialer might place five calls per available agent. If the answer rate drops to 10 percent, it dials more aggressively — maybe eight or ten per agent. If the answer rate spikes to 40 percent, it backs off to two or three per agent to avoid overwhelming the agent pool.

The target is to keep the abandon rate — calls that are answered by a live person but no agent is available to take the call — below 3 percent. The FCC requires that predictive dialers maintain an abandon rate at or below 3 percent measured over a 30-day period per campaign. Exceeding this threshold is a TCPA violation. The dialer’s algorithm constantly adjusts the dial ratio to maximize agent utilization while staying under the abandon ceiling.

In VICIdial specifically, this behavior is controlled through campaign settings: the dial level (auto or manual ratio), the adaptive dial settings (which use real-time drop percentage targets to automatically adjust pacing), and the hopper (the queue of leads ready to be dialed). VICIdial’s adaptive algorithm is one of the most configurable in the open-source dialer space — it can target a specific drop percentage and adjust the dial level every few seconds based on actual performance.

How predictive dialing works at the SIP level

Understanding what happens at the trunk level explains why carrier choice matters so much for predictive dialer performance.

When VICIdial (or any predictive dialer) decides to place a call, it sends a SIP INVITE through the configured trunk to the carrier. Each simultaneous call attempt is a separate SIP session — a separate INVITE, a separate call setup, a separate media stream if the call is answered. If the dialer is running 10 agents at a 3:1 ratio, there are 30 simultaneous SIP INVITEs in flight at any given moment. If the answer rate drops and the ratio increases to 5:1, there are 50 simultaneous sessions. A 100-agent floor at a 4:1 ratio generates 400 concurrent SIP sessions.

Each of those sessions requires the carrier to authenticate the trunk, validate the caller ID, perform STIR/SHAKEN signing, route the call, and handle the media. If the call is answered, an RTP audio stream is established. If the call goes to voicemail and the dialer detects it via AMD (Answering Machine Detection), the session is torn down. If the called number is disconnected, the carrier returns a SIP error response (typically 404 or 604) and the session ends.

The carrier’s capacity to handle this burst traffic without degradation determines your dialer’s effective performance ceiling. Here is where the problems show up:

Post-Dial Delay (PDD). This is the time between the SIP INVITE leaving your dialer and the called party’s phone starting to ring. On a good carrier, PDD is under 3 seconds. On a congested or poorly peered carrier, it can reach 5 to 8 seconds. In a predictive dialing environment, excessive PDD cascades through the algorithm — the dialer’s predictions about when agents will become free are based on expected call setup times. If PDD varies wildly, the algorithm makes bad predictions: too many calls connect at once (agents overwhelmed, abandon rate spikes) or too few connect (agents idle, efficiency drops).

Concurrent channel limits. Most SIP trunk resellers cap concurrent channels at 10, 25, or 50 per trunk. A 50-agent operation running a 4:1 ratio needs 200 concurrent channels. If the carrier caps at 50, the dialer cannot execute its algorithm — it gets SIP 503 (Service Unavailable) responses for every call beyond the cap, effectively throttling the operation to a fraction of its potential. On SIPNEX, concurrent channels are unlimited. Your capacity is your bandwidth and your server, not an artificial carrier-imposed limit.

STIR/SHAKEN attestation level. Every one of those SIP INVITEs gets signed by the carrier. If the carrier is a reseller passing calls upstream, the signing happens at the upstream carrier — at B-level, because the upstream carrier has no direct relationship with you. Every call in your campaign goes out with partial attestation. On carrier handsets with aggressive spam filtering (T-Mobile, AT&T), B-level calls get treated with more suspicion than A-level. Across 400 concurrent sessions, the attestation difference compounds into a measurable answer rate gap.

Codec and media quality. VICIdial recommends G.711u (ulaw) as the primary codec for maximum audio quality. Each G.711 call uses approximately 85 kbps of bandwidth. At 200 concurrent calls, that is 17 Mbps of sustained audio traffic in each direction. The carrier’s media infrastructure must handle this without packet loss or jitter. Cheap carriers with oversubscribed media gateways introduce audio quality problems that the dialer cannot compensate for — choppy audio, one-way audio, echo — all of which cause agents to lose calls and consumers to hang up.

Predictive vs progressive vs preview dialers

Predictive dialing is not the only mode, and it is not always the right one. Understanding the three primary modes helps you match the technology to the campaign.

Predictive dialing places multiple calls per available agent based on algorithmic predictions. Highest efficiency — agents spend the maximum percentage of their time talking to live humans. Highest compliance risk — the algorithm can miscalculate and cause abandoned calls. Best for: high-volume B2C campaigns with large lead lists, commodity offers, appointment setting, collections. VICIdial campaigns using adaptive dialing at a 3 to 5 percent drop target are the standard implementation.

Progressive dialing (sometimes called power dialing) places exactly one call per available agent. When an agent finishes a call and becomes available, the dialer places one new call for that agent. No algorithmic prediction, no multi-line dialing, no abandon rate risk from over-dialing. Lower efficiency than predictive — agents wait through each ring cycle — but zero risk of the abandon rate exceeding the FCC’s 3 percent threshold. Best for: regulated industries (healthcare, financial services), campaigns with compliance-sensitive leads, and any operation where a single abandon-rate violation could be catastrophic.

Preview dialing shows the agent the lead record before the call is placed. The agent reviews the information — name, account history, previous interactions, notes — and decides whether to dial or skip. The agent initiates the call manually. Lowest efficiency of the three modes because every call includes a human review step. But the highest quality conversations because the agent is prepared. Best for: complex sales (insurance, real estate, B2B), high-value leads where personalization matters, and any campaign where the conversion value justifies the lower dial volume.

Most sophisticated operations use all three modes across different campaigns. A collections shop might run predictive for right-party-contact campaigns (high volume, low touch) and preview for settlement negotiation calls (high value, preparation required). VICIdial supports all three modes and allows switching between them per campaign without changing the underlying trunk configuration.

Why your carrier matters more than your dialer software

This is the point most operators miss, and it is the reason we wrote this article from the carrier perspective rather than as a software comparison.

VICIdial is one of the most capable open-source predictive dialers available. It has been in active development for over 15 years. Its adaptive algorithm is well-tested. Its configuration options are extensive. It runs on commodity hardware. Thousands of call centers worldwide use it to dial millions of calls per day.

But VICIdial — or any dialer — is only as good as the SIP trunk feeding it. The dialer is the brain. The trunk is the spine. If the spine is broken, it does not matter how smart the brain is.

A carrier that throttles concurrent channels means your dialer cannot execute its algorithm. The adaptive pacing in VICIdial calculates the optimal number of simultaneous calls based on real-time metrics. If the carrier returns 503 errors because you hit a channel cap, the dialer’s predictions are invalid — it expected those calls to either connect or fail normally, not get blocked at the carrier level. The result is unpredictable agent utilization and inaccurate drop rate calculations.

A carrier that provides B-level attestation means your answer rates have a ceiling. VICIdial can optimize dial pacing, AMD detection, CID rotation, and lead prioritization. It cannot change the fact that every call from a B-level trunk hits the recipient’s phone with a lower trust signal than A-level. The answer rate penalty from B-level attestation — typically 10 to 20 percent relative to A-level on the same leads — is a carrier-layer problem that no amount of dialer optimization can fix.

A carrier with high or variable PDD means your algorithm’s predictions are wrong. VICIdial’s adaptive dialing assumes a relatively consistent call setup time. If PDD varies from 2 seconds to 8 seconds across calls because the carrier has inconsistent peering or congested signaling infrastructure, the dialer over-dials on some cycles (causing abandons) and under-dials on others (causing idle time). The algorithm cannot converge on optimal pacing because the underlying timing is unstable.

A carrier with poor media quality means your connected calls fail anyway. An agent connects to a live human, but the audio is choppy, there is echo, or one direction drops out entirely. The consumer says “hello” twice and hangs up. The agent marks the lead as a callback. The campaign metrics look worse than they should because the connected calls are not converting — not because the leads are bad or the agents are bad, but because the trunk is degrading the audio.

SIPNEX exists because we experienced every one of these problems as dialer operators before we became a carrier. We built the carrier we wished existed: unlimited channels, sub-3-second PDD, direct A-level STIR/SHAKEN attestation signed with our own SP-KI certificate, and media infrastructure sized for high-concurrency predictive dialing workloads. When you call our support about a VICIdial configuration issue, the person answering has run VICIdial campaigns themselves.

Choosing a carrier for your predictive dialer

Based on everything above, here is what to evaluate when selecting a carrier for your dialing operation.

Concurrent channel policy. Ask explicitly: is there a per-trunk or per-account concurrent channel limit? If the answer is anything other than “unlimited” or a number higher than your peak predicted concurrency, the carrier will throttle your operation. Most resellers cap at 10 to 50 channels. SIPNEX has no channel limit.

Post-dial delay. Ask for average PDD data. Under 3 seconds is acceptable. Under 2 seconds is good. If the carrier cannot tell you their PDD metrics, they probably do not monitor them, which means they do not optimize for them.

STIR/SHAKEN attestation. Ask the critical question: do you hold your own STIR/SHAKEN Service Provider certificate, or do you inherit attestation from an upstream carrier? If they inherit, you are getting B-level. If they hold their own cert and verify your DIDs, you are getting A-level. This is the single highest-impact factor for answer rates that is entirely within your control as a carrier selection decision.

Codec support. Confirm G.711u (ulaw) support as primary. G.729 as fallback. Opus if your platform supports it. Ask about transcoding — does the carrier transcode between codecs, or does it pass through natively? Transcoding adds latency and can degrade quality.

Pricing transparency. If you cannot see rates before talking to sales, you are paying for their sales team. Wholesale per-minute rates for high-volume predictive dialing should be in the $0.005 to $0.015 range depending on volume and destination mix. Per-channel fees are a red flag — they indicate the carrier is monetizing an artificial scarcity rather than charging for actual usage.

Support depth. Call the support number and ask a VICIdial-specific question: what codec and DTMF mode do you recommend for VICIdial AMD detection? If they cannot answer without escalating, their support team does not understand your workload.

Frequently asked questions

What is a predictive dialer and how does it work?

A predictive dialer is call center software that places outbound calls automatically, using statistical algorithms to predict when agents will become available and dialing multiple numbers in advance so that a live connection is ready the moment an agent finishes their current call. The dialer monitors answer rates, call durations, and agent availability in real time, continuously adjusting the number of simultaneous calls to maximize agent talk time while keeping the abandon rate below the FCC’s 3 percent threshold. The result is that agents spend 85 to 95 percent of their time in conversation rather than the 30 to 40 percent typical of manual dialing.

How does a predictive dialer differ from an auto dialer?

An auto dialer (also called a power dialer or progressive dialer) places one call per available agent — when an agent becomes free, the system dials one number for that agent and waits for the result. A predictive dialer places multiple calls per available agent based on algorithmic predictions about answer rates and call durations. The predictive approach is significantly more efficient because it accounts for the fact that most outbound calls are not answered — it dials enough numbers simultaneously that a live answer is statistically likely to coincide with an agent becoming free. The tradeoff is complexity and compliance risk: predictive dialing can produce abandoned calls if the algorithm over-dials, which is why the FCC caps the abandon rate at 3 percent.

Is predictive dialing legal under TCPA?

Predictive dialing itself is not illegal, but it creates TCPA obligations. If your predictive dialer calls cell phones — and in 2026, the vast majority of consumer phone numbers are cell phones — you need prior express consent for informational calls and prior express written consent for telemarketing calls. You must also maintain the abandon rate at or below 3 percent measured over a 30-day period per campaign, honor the National Do Not Call Registry and company-specific DNC requests, and restrict calling to 8:00 a.m. to 9:00 p.m. in the recipient’s local time zone. The legal question of whether a predictive dialer qualifies as an ATDS under TCPA was narrowed by the Supreme Court’s 2021 Facebook v. Duguid decision, but the safest approach is to treat your dialer as an ATDS and meet the higher consent standard regardless. See our TCPA compliance checklist for the full breakdown.

What kind of SIP trunk carrier do I need for a predictive dialer?

You need a carrier that provides unlimited concurrent channels (predictive dialers burst to high concurrency and channel caps throttle the algorithm), low post-dial delay under 3 seconds (variable PDD breaks the dialer’s pacing predictions), direct A-level STIR/SHAKEN attestation from a carrier that holds its own SP-KI certificate (B-level from a reseller costs you 10 to 20 percent answer rate), transparent per-minute pricing without per-channel fees, and support engineers who understand predictive dialer workloads. Most SIP trunk resellers fail on at least two of these requirements because they inherit their network characteristics from an upstream carrier they do not control. A direct carrier like SIPNEX is built specifically for this workload.

How many concurrent channels does VICIdial need?

The formula is: number of agents multiplied by the dial ratio equals required concurrent channels. If you have 50 agents running a 4:1 predictive ratio, you need 200 concurrent channels. If you have 100 agents at 3:1, you need 300 channels. During campaign ramp-up or when answer rates drop suddenly (triggering more aggressive dialing), the burst requirement can be 20 to 30 percent higher than the steady-state calculation. On a carrier with channel caps, you need to purchase a channel count that covers the burst scenario — which means paying for capacity you only use intermittently. On SIPNEX, channels are unlimited, so the calculation becomes purely about bandwidth: each G.711 call uses approximately 85 kbps, and you need enough upstream and downstream bandwidth to sustain your peak concurrent call count plus overhead for signaling and other traffic.

---

SIPNEX is an FCC-licensed carrier built by operators who run VICIdial themselves. Unlimited channels, direct A-level STIR/SHAKEN attestation, sub-3-second PDD, and support from engineers who know what sip.conf is. Request a dialer-grade trunk or see our published rates.