There are three ways a power dialer can put calls in front of your reps, and picking the wrong one does one of two things: it wastes your agents' time waiting between calls, or it burns through your phone numbers until the carriers start slapping "Spam Likely" on every one of them. Most teams pick based on a demo where the vendor showed off the fastest, flashiest mode — and then wonder six weeks later why their answer rate cratered.
Full disclosure: I work for Ready, and we ship a Power Dialer. So I have a horse in this race. But the mode you choose matters more than the vendor you choose, and I'd rather you run the right mode on someone else's tool than the wrong one on ours. Here's the actual mechanics.
The three modes, in one sentence each
- Preview — the system shows the rep the contact record, the rep decides when to click dial. One call at a time. Human in the loop before every dial.
- Progressive — the system auto-dials the next number the moment a rep frees up. One line per rep, no waiting for a click, but no dialing ahead either.
- Predictive — the system dials multiple numbers per available rep, betting that most won't answer, and connects whoever picks up to a free agent. This is where the trouble lives.
The difference that matters isn't speed on paper. It's the ratio of dials to available agents, because that ratio is exactly what determines how many live humans pick up a phone and hear dead air.
Preview: for high-value lists and complex sales
Preview dialing is the slowest mode and, for the right list, the smartest. The rep sees who they're about to call — the notes, the last touch, the deal size — and dials when ready. No dial happens without a person deciding to make it.
Use it when:
- Each contact is worth real money (six-figure B2B, listings, high-ticket services).
- Reps need 30–60 seconds of context before every call.
- Your list is small enough that connect rate matters more than dial volume.
The tradeoff is obvious: talk time per hour is low because reps spend time reading before dialing. If a rep makes 40 preview dials in an hour and connects on 8, that's fine when each connect is a $20k pipeline opportunity. It's a disaster when you're working a list of 5,000 and need volume.
Cost-wise, preview is cheap on minutes because you're not burning connect time on dead air. On Ready's Power Dialer, minutes bill in 6-second increments at $0.05/min on Pro or $0.0375/min on Team — so a slow, deliberate preview cadence keeps your minute bill low. See the Power Dialer pricing for the per-agent breakdown.
Progressive: the default that almost everyone should run
Progressive dialing is one line per rep, dialed automatically. The rep hangs up, the next number fires. No click, no dial-ahead, no gambling on who's available.
This is the mode most teams should default to, and here's why: because it's one dial per free agent, your abandonment rate is structurally near zero. There's always an agent waiting for the connect, because the dial only happened because an agent was free. You get most of the throughput gain over preview (no manual clicking, no reading delay if your reps don't need it) without the risk that predictive introduces.
Progressive fits:
- Teams of any size working warm or semi-warm lists.
- Speed-to-lead workflows where a new inbound lead needs a dial now. Ready's Team plan pairs progressive-style auto-dial with an instant SMS the moment a lead comes in — the compressed speed-to-lead cadence that beats six calls over three days.
- Anyone who wants more talk time than preview without touching predictive's abandonment risk.
If you're not sure which mode to run, run progressive. You'll almost never regret it.
Predictive: fast, and the fastest way to torch your numbers
Predictive is the mode vendors love to demo because the talk-time numbers look incredible. The system dials, say, 2–3 numbers per available agent because it predicts most won't answer. When one does, it routes to a free rep. The math assumes a certain answer rate and paces dials to keep agents constantly talking.
The problem is what happens when the prediction is wrong — when more people answer than there are free agents. Those extra live humans pick up and hear... nothing. A pause, then a click. That's an abandoned call, and it's not just rude. It's tracked, regulated, and it's how carriers decide your numbers are spam.
The 3% cap isn't a suggestion
TCPA guidance holds telemarketing abandonment to roughly 3% of answered calls, measured over a 30-day campaign. Blow past it and you're exposed on two fronts:
- Regulatory — abandoned-call violations stack up per call.
- Carrier reputation — high abandonment and short-duration calls are exactly the signals carriers' analytics use to flag a number. Once flagged, your answer rate drops because "Spam Likely" shows on the recipient's screen, which makes you dial harder, which makes it worse.
We went deep on the math in the 3% abandonment cap post — the short version is that predictive dialing at an aggressive line ratio hangs up on roughly 1 in 12 answered calls if you're not careful. The line-ratio math here shows exactly where the ratio breaks.
Predictive only makes sense with enough agents to smooth out the variance. Three agents can't run predictive safely — one lucky answer streak and everyone's on a call while two live humans get abandoned. You need a pod of 8–10+ agents before the statistics work in your favor, plus real-time pacing that throttles down when abandonment climbs.
Matching mode to your situation
| Preview | Progressive | Predictive | |
|---|---|---|---|
| Dials per free agent | 1 (manual) | 1 (auto) | 2–3+ (auto) |
| Abandonment risk | None | Near zero | High if mispaced |
| Best team size | Any | Any | 8–10+ agents |
| Best list quality | High-value, warm | Warm to semi-warm | Large cold/semi-cold |
| Talk time per rep | Low | Medium | High (when safe) |
| Number-flagging risk | Low | Low | High |
A few honest calls:
- Solo rep or small team? Preview or progressive. Predictive with fewer than ~8 agents is a coin flip against your own numbers.
- High-ticket, relationship sale? Preview. The context is worth the slower pace.
- Big list, big floor? Predictive can work — but only with abandonment monitoring and a conservative line ratio. Start at 1.5 lines per agent, not 3.
- Everyone else? Progressive. Boring, safe, effective.
Protecting your numbers regardless of mode
Mode is only half the number-reputation story. Even perfect progressive dialing gets flagged if the underlying list is dirty or your numbers get overused.
- Scrub before you dial. A TCPA and DNC litigator scrub at $0.005/contact suppresses the numbers most likely to file a complaint. Scrubbing 100,000 contacts runs $500 — versus $500–$1,500 per text or call for a single TCPA complaint.
- Don't over-rotate local numbers. Cycling local-presence numbers to lift answer rates can get all of them flagged in a week if you spread volume too thin across too many numbers.
- Respect quiet hours. Dialing at the wrong local time drives both complaints and wasted dials.
Ready's Power Dialer ships call recording, voicemail drop, transfer/barge/whisper, and speed-to-lead auto-dial on the Team plan — but the compliance plumbing (scrub, quiet hours, opt-out handling) is what keeps your numbers alive long enough to use any of it.
The practical takeaway
Predictive dialing isn't evil — it's a tool with a narrow safe operating range that most teams run wildly outside of. If you have a large floor, real pacing controls, and you watch abandonment like a hawk, it'll drive more conversations per hour than anything else. If you don't, it'll quietly kill your phone numbers and hand you a compliance liability.
Progressive is the mode I'd point 90% of teams at. Preview for the high-value, low-volume work. Predictive only when the agent count and the monitoring justify it.
If you want to sketch the minute math for your team size, the Power Dialer plans lay out the per-agent and per-minute costs, and you can start free — the Free tier includes 1 agent, 1 number, and 500 minutes a month to test a cadence before you commit a floor to it.