For the Five9 admin staring at a silent outbound floor: an ordered troubleshooting tree, where each check lives in the VCC admin, and the fix for every one. Written by operators who run these campaigns daily.
Why is my Five9 outbound campaign not dialing?
Usually one of five things. The campaign is stopped or in the wrong dialing mode; the schedule or timezone rules are blocking every record; the lists are spent, filtered to zero, or parked on redial timers; no agents are READY with the campaign's skill, which stalls pacing; or DNC and ANI configuration is suppressing dials. Work the checks in that order and most faults surface inside the first three.
Read what the dialer is already telling you before you change anything.
An outbound campaign that has gone quiet usually announces the reason, if you read the panel before touching it. Open the VCC Administrator application, find the campaign under Campaigns, and look at its state. We run Five9 floors daily, and the number of urgent escalations that turn out to be a campaign nobody restarted after maintenance, or one a supervisor stopped on Friday and forgot, is genuinely humbling.
| Campaign state | What it usually means | First move |
|---|---|---|
| Not Running | Stopped manually, stopped by its schedule, or never started after an edit | Start it, then check the Schedule tab before blaming a colleague |
| Running, but silent | A pacing, agent, suppression, or capacity fault further down this tree | Work the checks below in order |
| Out of Numbers | Every attached record is dialed to its attempt cap, filtered out, suppressed, or parked on a redial timer | Go straight to the list and suppression checks |
Confirm the dialing mode while you are there. A campaign flipped to preview for a test sits waiting for agents to click through records rather than auto-dialing, and on a busy panel that looks identical to a dead campaign. Then restart deliberately. Plenty of property changes only apply after a stop and start, so an edited campaign that has not dialed since the edit needs a clean restart before you conclude anything else. Five9's own outbound campaign troubleshooting page leans on the same basic checks, which tells you how often they are the answer.
The dialer is often doing exactly what the calendar told it to.
The second most common fault is a calendar. Every outbound campaign has a schedule, the days and hours it may run, and separately a set of timezone dialing rules that decide when each individual record is legal to call in its own local time. Either one can silence a campaign that looks perfectly healthy.
The symptom has a signature. A campaign that dials happily in the afternoon but sits silent all morning is almost always waiting for records to enter their local window. Load a California-heavy list and start dialing at 9 a.m. Eastern, and nothing is dialable for three hours. The dialer is not broken; it is refusing to call people at 6 a.m., which is the behavior you want. The federal default under 47 CFR 64.1200(c)(1) prohibits telephone solicitations before 8 a.m. or after 9 p.m. local time at the called party's location, and several states have tightened their own windows and added day restrictions in recent years. Rules in this space keep moving, so check the current text for the states you dial before widening any window, and take advice from qualified counsel where the exposure justifies it.
Open the campaign's Schedule tab for the run window, then the time zone dialing options for how Five9 decides each record's local time. Check which data drives it: records get a timezone from area code and postal data, so a list imported with mangled zip codes can put half your records in the wrong window. If you dial mobiles and landlines on different rules, check the per-number dialing hours as well. The fix is rarely to widen the window. It is to sequence your lists so the floor always has something legal to dial: start Eastern, follow the sun west.
Our free Five9 Health Check is 17 questions and takes about ten minutes. It flags the configuration gaps that cause exactly this kind of morning, before the morning happens.
Three different list faults produce the same silence.
When the state reads Out of Numbers, or the campaign runs dry mid-shift, the list machinery is the place to look. A list not dialing is almost always a list that is spent, filtered, or broken on import, and Five9's own admin guidance on the Out of Numbers state names much the same culprits we find on floors.
The filter case deserves particular paranoia because it produces no error anywhere. The campaign runs, the list is attached, and zero records qualify. When we audit a silent campaign, the profile filter is the first thing we screenshot. Check it against real field values exported from the list, not against what the values are supposed to look like.
Automatic modes pace off ready agents, and zero ready agents means zero dials.
Power, progressive, and predictive modes all pace against agents who are ready to take a connect. No agents in a READY state carrying the campaign's skill means the dialer places nothing at all. It is not stalled; it is correctly declining to dial calls nobody could answer.
Check three things in the supervisor view. Agents are logged in. They are READY for calls, rather than Not Ready or parked in After Call Work. And they actually carry the skill the campaign delivers to, because a skill mapping removed during a tidy-up produces a floor full of ready agents the campaign cannot see. User-to-skill assignments are the first casualty of every reorganization we audit.
Predictive mode adds a subtlety. Its pacing algorithm needs a statistically useful pool to predict against, and Five9's own campaign documentation advises against running predictive with fewer than 15 consistently active agents. Run it with four and pacing turns glacial or stalls entirely, which reads as a dead campaign from the floor. The honest answer for small teams is power or progressive mode, not a fight with the predictive math. We cover the tuning side in our predictive dialer settings guide.
Pacing throttles produce near-silence too. A power campaign set to one call per agent with aggressive answering machine detection feels dead between connects. And a predictive campaign held to a tight abandonment ceiling will slow itself down to stay under it, because the dialer treats that cap as a hard constraint. The cap exists for a reason: US federal rules limit abandoned calls, counted as abandoned calls divided by calls answered by a live person, to 3 percent measured over a 30-day period for a single campaign. The dialer protects that number by dialing less. If pacing against the abandonment cap is the real problem, our piece on abandonment rules and predictive dialing goes deeper.
DNC lists, scrubs, and redial timers all fail silent by design.
Suppression logic removes records before they ever reach the dialer, and none of it raises an error. Start with the domain Do Not Call list: it sits above your campaigns and its size only ever grows. Then look at dispositions, because dispositions are where suppressions breed. Any disposition can be configured to add the number to the DNC list, and we have audited more than one floor where a routine wrap-up code had that box ticked, quietly consuming the database one call at a time. If your dialable pool shrinks faster than your dial count explains, that is the signature.
Third-party scrubs, litigator lists, reassigned-number checks, and consent expiry usually arrive as list filters or pre-import processing, so they show up in the filtered-to-zero check above rather than anywhere labeled DNC.
Redial timers are the sneakiest version. Every disposition carries a retry timer and an attempt cap: No Answer might retry in four hours, Busy in thirty minutes. Set long timers across your main dispositions and a modest list is fully parked by late morning, every remaining record waiting out a timer. The campaign shows Running, places calls in bursts as timers expire, and sits silent in between. Open the campaign's disposition settings and read the timers against your list size and daily attempt plan. That arithmetic is rarely done and always revealing.
No valid ANI means nothing lawful to present. No free ports means nowhere to dial from.
A campaign with no valid outbound ANI has no caller ID to present, and depending on configuration it will fail dials immediately or decline to place them. The ANI lives on the campaign profile, so a campaign moved to a different profile during a reshuffle can lose its caller ID without anyone touching the campaign itself. Open the profile's ANI settings and confirm a valid number is assigned; the domain-level ANI management screen shows what exists to choose from. And remember that an ANI which dials fine but displays as Spam Likely is a different failure with similar revenue consequences. We wrote up how to fix flagged caller ID separately.
Capacity faults look different: dials trickle at peak and flow off-peak. Your Five9 domain has finite concurrent call capacity shared across everything it does, inbound included, and a hungry autodial campaign or an inbound spike can starve an outbound campaign of ports at exactly the hours you care about. The giveaway is a time-of-day pattern that matches no schedule and no timezone rule. This one is worth confirming with Five9 support directly, because provisioned capacity is theirs to state, and the fix may be commercial rather than technical.
Nothing in the silent campaign is wrong. Look at its neighbors.
Run several outbound campaigns against a shared agent pool and the dialer has to decide who gets the next ready agent. Five9 evaluates campaigns by priority on each dialing cycle, highest first, with ratio settings available to split capacity proportionally instead. This is documented and sensible, and it is also the source of a very specific fault: the low-priority campaign that never dials while its high-priority sibling is awake, then springs to life at 6 p.m. when the sibling's schedule closes.
Nobody diagnoses this from inside the silent campaign, because nothing in it is misconfigured. Check what else shares its agents. If a higher-priority campaign paces hungrily enough to absorb every ready agent, the campaign below it starves indefinitely, and the tell is that it dials exactly when the other one stops. The fix is a deliberate choice between priority, ratio, or the blend of both that Five9 supports, made against what the business actually wants dialed first rather than whichever campaign an admin happened to create first. While you are in multi-campaign territory, compare profile filters across campaigns too. Two campaigns sharing a list with complementary filters can each assume the other is dialing the records that, in fact, neither is.
Five9 support and a partner solve different halves of this problem.
If you have worked the whole tree and the campaign still will not dial, escalate with evidence rather than adjectives. Campaign name, exact timestamps, a handful of example records that should have dialed, a screenshot of the state, and the checks you have already ruled out. That package cuts resolution time dramatically whoever you send it to.
| Symptom | Best first call |
|---|---|
| Campaign stuck in a state that survives a clean stop and restart | Five9 support |
| Telephony faults: dead air, failed call setup, carrier-level errors | Five9 support |
| Domain capacity, port limits, ANI provisioning | Five9 support |
| Filters, pacing, schedules, dispositions, list strategy | A partner or experienced operator |
| Campaign dials, but results are poor | A partner, and honestly, a strategy conversation |
The distinction that matters is this. Five9 support is good at establishing that the platform is doing what your configuration tells it to do. They will not tell you the configuration is a bad idea, that your redial timers guarantee a parked list by mid-morning, or that predictive was the wrong mode for nine agents. That judgment is an operator skill, and it is the gap a partner fills; it is also what our managed services clients stop having to think about at 8:55 on a Monday. Whoever you call, call them with the tree above already worked. The expensive part of this fault is never the fix. It is the day of silence before someone found it.
It means the campaign has no dialable records left. Everything attached is either dialed to its attempt cap, excluded by the campaign profile's list filter, suppressed by DNC, or parked on a disposition redial timer. It is a state, not an error. Check list penetration first, then the profile filter, then redial timers. If the list is genuinely worked to completion, the campaign has finished and needs fresh data rather than a restart.
Confirm the agents are READY with the specific skill the campaign delivers to, not merely logged in. Then check pool size: predictive pacing needs a meaningful number of consistently active agents, and Five9's documentation advises against using it with fewer than 15. Small teams should run power or progressive mode instead. Finally, check whether a tight abandonment ceiling is throttling the pace, because the dialer slows itself to stay under that cap.
Look at the campaign's attached lists in the VCC Administrator and run the list penetration reporting to see how much of each list has been dialed to completion under your attempt rules. A list showing high penetration has been worked, not lost. Resetting a list re-opens spent records for dialing, but treat that as a deliberate decision with compliance weight: re-dialing exhausted records is sometimes right and often not.
Treat a stop and start as mandatory after edits. Some property changes apply to a running campaign, but plenty only take effect on restart, and the failure mode is silent: the campaign keeps running on the old configuration and nothing tells you otherwise. Our habit on the floors we run is to make the change, stop the campaign, start it again, and watch the first dialing cycle before walking away.
That is the signature of disposition redial timers. Every disposition carries a retry timer and an attempt cap, so once the fresh records are worked, everything remaining is parked waiting out a timer. The campaign bursts back to life as timers expire and sits silent in between. Read your main disposition timers against list size and daily attempt plan; long timers on No Answer and Busy can park a modest list by late morning.
Yes, and it does so without raising an error anywhere. The campaign profile's list filter applies to every attached list, and a criterion that matches zero records produces a running campaign that dials nothing. Common causes are a field the new list never populates, a value format mismatch such as FL against Florida, or a stale date range. Check the filter against real field values exported from the list, not against what the values should look like.
The federal rule at 47 CFR 64.1200(c)(1) prohibits telephone solicitations before 8 a.m. or after 9 p.m. local time at the called party's location, and Five9's timezone dialing features enforce windows per record. Several states set tighter windows and additional day restrictions, and these have changed in recent years, so verify the current state rules for the lists you dial and take qualified legal advice where the exposure matters.
Call Five9 support for platform faults: a campaign stuck in a state that survives a clean restart, telephony errors, dead air, ANI provisioning, or domain capacity questions. Configuration and strategy questions, meaning pacing, filters, dispositions, list design, and multi-campaign priorities, are operator territory. Support will confirm the platform works as configured but will not judge whether the configuration is wise. Bring evidence either way: campaign name, timestamps, and example records.
Send us the campaign name and what you have already checked. We run Five9 outbound floors every day, and most of these faults take an experienced pair of eyes minutes to find.
Talk to us