RMT Engineering Logo
Outbound campaigns and dialing

Predictive Dialer

Three dialing modes on one console, paced under a hard abandon-rate cap the engine throttles to stay below, not a number it aims at. Every contact clears do-not-call, consent and its own calling window before the number is dialled. Not in a report next Tuesday.

  • Hard abandon-rate cap
  • Checks run before the dial
  • The contact's time zone, not yours
We reply within one business day. No newsletter.
3 Dialing modes on one console
DNC Checked before every dial
Time zone And holiday-aware calling windows
Hard cap Paced under the abandon cap

Dial at pace, and hold back the calls that should never be placed

Campaigns, lists, scripts, retry policies and reporting sit in one console. The mode a campaign dials in is a setting on the campaign, not a separate product with its own licence. Outbound shares a tenancy with your inbound channels, so a customer who rings back an hour later reaches an agent looking at the same record, with the attempt already on it. Underneath all of it sits one compliance gate. No number is dialled until the contact has cleared it.

Outbound that dials fast and stays compliant

Most outbound teams end up choosing. Dial hard and the abandon rate climbs, someone gets called at nine in the evening local time, and the compliance report lands a fortnight later listing calls nobody can unmake. Slow the dialer down and half the shift goes on ring tone. Here the pacing engine and the compliance gate are the same product. The gate blocks the call instead of logging a warning about it, so you can push speed against the cap without pushing against the rules.

How it fits together

The path a conversation takes through Predictive Dialer

How the compliance gate decides: every contact is checked against do-not-call lists, captured consent, the contact's own calling window and two-party consent requirements before a call is placed. A failed check blocks the call instead of recording a warning afterwards. Contact from the list Compliance gate — every call, before it is placed Do-not-call lists National, internal and client-supplied Consent Captured, verified and dated Calling window The contact's timezone and public holidays Two-party consent announcement Generated where the jurisdiction requires it All checks pass The call is placed Any check fails Blocked, and the reason logged

The gate runs before the dial, not as a report afterwards. A contact that fails any check is held with the reason recorded, so an audit shows what was blocked and why.

By the numbers

Few Facts - at a glance

  • 3

    dialing modes, switched per campaign or per hour

  • DNC

    enforcement built in, checked before the dial rather than audited after it

  • Time zone

    and public-holiday aware windows, read from the contact's own location

The process

How It Works

The four stages, end to end

Four things happen between a list arriving and an outcome being written back. The second one is what auditors ask about. The compliance gate sits between the audience and the dial, so a contact that fails a check is never called at all. Below is what the gate checks, and what happens on each branch.

  1. Build the audience

    AND/OR rules run over CRM records and imported CSV files. Duplicates get detected and dropped. Each contact carries a lead score out of 100 from a rule engine you can read line by line, and the audience size is on the screen before you commit to it.

  2. Clear the compliance gate

    Every contact is checked against national, internal and client-supplied do-not-call lists. Then against the consent you hold for it. Then against its own calling window, read from the contact's time zone and local public holidays rather than from your office clock. A contact that fails any of those is held, with the reason recorded against it.

  3. Dial at the set pace

    Predictive, power or progressive, set by the campaign and by the hour. Pacing throttles as the abandon rate climbs toward its cap. Concurrent-call limits hold. Fair-queue scheduling stops one big campaign from starving the smaller ones running beside it.

  4. Disposition, retry, sync

    The outcome is written on the contact record and pushed back to your CRM idempotently on the call ID, so a write retried after a dropped connection updates the row it already made instead of creating a second one. Anything short of a connect schedules another attempt. Backoff and answer history decide when.

Features

Every capability you need in one module

1. Three modes, switched per campaign or per hour

Predictive dialing paces ahead of the agents against predicted answer rates, held under a hard abandon-rate cap. Power places a fixed ratio of calls per available agent, two to one if that is what you set. Progressive places one call per agent and waits, which is why it cannot abandon a call at all. Most floors run all three across a day. Predictive through the morning, power after one, progressive from four.

  • A hard abandon-rate cap, not an aspiration
  • Fair-queue scheduling across campaigns
  • Concurrent-call limits, set per campaign
  • Voice, SMS and WhatsApp campaigns, with inbound replies landing on the same contact record

See the AI voice agent that places the call

Illustrative interface. Campaign names, ratios and schedules are sample data, not a performance claim.

2. Retries that learn the contact

Backoff runs 1 hour, then 4 hours, then a day, then three days, then a week. That is the default. Attempt caps are set per campaign, and public holidays are skipped entirely. Best-time-to-call is read from your own answer history, so the fourth attempt lands in the hour this contact has picked up before, not in the hour your list happened to start.

  • Backoff of 1 hour, 4 hours, 1 day, 3 days and 1 week
  • Per-campaign attempt caps
  • Public holidays skipped, in the contact's own jurisdiction
  • Best-time-to-call read from your own answer data, not a rule of thumb

Bring a list and a script

We will load a sample list, run the compliance gate over it and dial a live test campaign with you watching.

Book a demo with our team

Illustrative interface. Attempt outcomes, timings and answer-likelihood curves are sample data, not a performance claim.

3. One script, every contact

A template engine with variables, conditionals and loops, plus a catalogue of every variable on the record and a live preview rendered against a real contact before anything goes out. The AI agent adapts the wording as it speaks. Your own agent reads the same rendered script in the Agent Desk. One template, maintained once.

  • Variables, conditionals and loops
  • A catalogue of every variable available on this record, so nobody guesses at a field name
  • Live preview against a real contact before it goes out
  • The same script in the AI agent and in the Agent Desk

See the Agent Desk your team reads it in

Illustrative interface. Variable names, contact details and script text are sample data, not a performance claim.

4. Lead scores you can explain out loud

AND/OR filter rules run over demographic and behavioural signals, and CSV import catches duplicates on the way in. Scoring is 0–100 on a rule engine you can read, with BANT, CHAMP and MEDDIC presets to start from. So when a sales manager asks why a contact scored 74, the answer is the rules that fired. Not a shrug.

  • AND/OR rules over demographic and behavioural signals
  • CSV import with duplicate detection
  • Explainable 0–100 scoring, with BANT, CHAMP and MEDDIC presets if you would rather not start from a blank rule set
  • Results synced back to Salesforce, HubSpot, Zoho or Pipedrive, idempotent on call ID

See the workflows that fire after a disposition

Illustrative interface. Rules, audience sizes, names and scores are sample data, not a performance claim.

5. Per campaign, and across all of them

Outcome breakdown, connect and conversion rates, blocked-by-gate counts with the reason attached, per-agent performance, drop-off by hour, best-time-to-call. All of it runs on the same reporting stack as every inbound channel, so an outbound connect rate and a floor occupancy figure sit in one view rather than in two exports stitched together on somebody's laptop the night before the meeting. Blocked counts sit beside connects. One screen answers the compliance question and the performance question.

  • Outcome breakdown, connect rate and conversion rate
  • Blocked by the gate, with the check that failed
  • Per-agent performance, on the same numbers the inbound floor is measured on
  • Drop-off by hour

See the 76 reports behind these numbers

Illustrative interface. Volumes, rates and outcome counts are sample data, not a performance claim.

Use Cases

Where Predictive Dialer delivers value

Compliance gate and calling windows

A survey campaign that holds itself back

A national utility (illustrative)

Scenario

A post-resolution survey list spans four time zones. Contacts whose local clock sits outside the 10:00 to 19:00 window are held rather than dialled. A regional public holiday takes a whole state out of the run. Eighteen numbers on a client-supplied suppression file never reach the dialer at all.

Outcome

The campaign shows as held instead of quietly ringing people at the wrong hour, and every blocked contact carries the reason it was blocked, so the compliance review is a filter on a report rather than a reconstruction from call logs three weeks later.

Exponential backoff and best-time-to-call

The fourth attempt is the one that connects

A mid-market lender (illustrative)

Scenario

A collections list runs on a five-attempt cap. The first three land on no answer and a busy tone, spaced an hour, four hours and a day apart. The fourth is scheduled into the early-evening window where this contact has answered before. It connects.

Outcome

Attempts get spent where the answer history says they are worth spending. And a contact who never picks up in the morning is not called four mornings running.

Shared scripts and CRM write-back

Renewals worked by people and an AI agent at once

A B2B software reseller (illustrative)

Scenario

A renewals audience is split down the middle. The AI voice agent takes the straightforward accounts in progressive mode, and your team takes the ones with an open complaint behind them. Both work from the same script template, with the loyalty-rate branch firing for anyone past two years, and both dispositions write back to the CRM on the call ID.

Outcome

One script to maintain instead of two that have drifted apart by November. No duplicate records when a call is retried either, because the write-back is keyed on the call ID.

At a glance

Specification

The numbers and limits, without the sales copy

Specification for Predictive Dialer
Specification Detail
Modes Predictive (paced on agent availability and predicted answer rate), power (fixed ratio per agent), progressive (one call per agent)
Pacing Hard abandon-rate cap, concurrent-call limits, fair-queue scheduling across campaigns
Compliance DNC lists (national, internal, client-supplied), consent capture and verification, two-party-consent announcements, calling windows
Windows IANA time zone and public-holiday aware, enforced per contact
Retries Exponential backoff at 1 hour, 4 hours, 1 day, 3 days and 1 week, with per-campaign attempt caps and holidays skipped
Audiences AND/OR filter rules, CSV import with duplicate detection, 0–100 explainable scoring with BANT, CHAMP and MEDDIC presets
Scripting Variables, conditionals, loops, variable catalogue, live preview against a real record
Channels Voice, SMS, WhatsApp campaigns with inbound reply handling
CRM sync Salesforce, HubSpot, Zoho, Pipedrive, idempotent on call ID
Reporting Outcome breakdown, connect and conversion rate, blocked-by-gate counts, per-agent performance, connect rate by hour
FAQ

Questions,
answered

What teams ask us before they roll out Predictive Dialer — how it works, what it needs from your side, and what happens when it gets something wrong

Still not sure?

Talk to a specialist and get a straight answer.

Ask our team

The OptiML dialer builds audiences from CRM sync and CSV import, with AND/OR filter rules and duplicate detection running over both. Outcomes are written back to Salesforce, HubSpot, Zoho or Pipedrive, keyed on the call ID so that retrying a failed write cannot leave you holding two records of one conversation. Nothing has to be exported and re-imported by hand.

The OptiML dialer needs your internal suppression list, any do-not-call files your clients supply, and whatever consent records you already hold. That is the part only you have. National DNC checking and calling-window rules ship with the module, and windows are enforced per contact from IANA time zones and local public holidays. Where a jurisdiction requires a two-party consent announcement, the dialer generates it.

The call is not placed. The OptiML dialer holds that contact on the campaign with the failed check recorded against it, so an audit shows what was blocked and why rather than showing a call that should never have happened in the first place. Blocked counts appear in campaign analytics beside connects and conversions, with the failed check attached to each one.

Yes. Human agents and an OptiML AI voice agent can work the same campaign from the same script template, so the variables, conditionals and loops are maintained in one place instead of two that drift. The AI agent speaks it naturally and adjusts phrasing as the conversation moves. Your agent sees the same script rendered in the Agent Desk, with the customer record beside it.

The OptiML dialer is a module of the OptiML CX Platform, deployed in the region you choose and on the same tenancy as your inbound channels. Contact data stays inside that deployment. The only outbound traffic is to the telephony, messaging and CRM providers you connect, so the list of third parties touching your contacts is a list you set. Talk to us about your carrier and your CRM before you plan a rollout.

Keep exploring

Additional supporting modules

Native modules that share the same record, knowledge and governance

Platform module

Voice AI Agents

The agent on the other end of the dial. Barge-in, mid-call language switching and provider failover.

Explore
Platform module

Automation & Workflows

What happens after the outcome is written: CRM updates, follow-up messages, approvals and escalations.

Explore
Platform module

Analytics & Reporting

Campaign numbers land in the same reports as every inbound channel. 76 reports, six export formats.

Explore
Connected solutions

Where Predictive Dialer is used

The Solutions pages that lean on this module, and what it looks like once it is configured for a particular floor, job title or job to be done.

16 solutions built on this module

By industry

Sector floors that run on it

9

By role

Job titles that answer for it

2

By use case

Jobs it is put to work on

5
Browse every solution
Talk to a specialist

Bring a list, and we will dial it with you watching

Send a sample contact list and one of your outbound scripts. We will build the audience, run the compliance gate over it so you can see what gets held and why, and dial a live test campaign while you watch. Your list, your script.

  • 30 minutes
  • Your own list and script
  • No slide deck unless you want one

Book your slot

Leave your email and our team will come back to you within one business day.

or reach us directly
Call our team Email us

Your details stay private. We never share them.

This website uses cookies.

Cookies are small text files that allow us to create the best browsing experience for you on our site. By continuing to use this website or clicking "Accept & Close", you are agreeing to our use of cookies. To understand how we use cookies or how to manage them, please see our cookies policy.

Ask OptiML

Powered by RMT Engineering