Predictive Dialer Abandonment Spike Before a Capacity Crunch
Ridgeway Connect runs a 60-seat outbound floor selling residential internet upgrades. The predictive dialer is set to 2.1 lines per available agent; average talk time is 5:30, after-call work 70 seconds, and 12% of dials reach a human. Over the last 30 days the campaign's abandonment rate is 6.4%, with spikes between 09:00 and 11:00, when eight agents are in new-hire training and logged out of the dialer. A prospect complained that she was called four times in one day and abandoned twice. The dialer vendor reports abandonment using a 2-second answer threshold, and the operations manager has been reporting the number weekly by dividing total abandoned calls by total answered calls across all campaigns, including a small B2B campaign. Sales leadership wants lines-per-agent raised to 2.6 to offset a 9% month-over-month drop in connects. The compliance team has not reviewed dialer settings since January.
1. Diagnose why abandonment is running at 6.4% and specify what you would change before the next shift.
Read the worked answer
Separate measurement error from conduct first. The 6.4% pools all campaigns, including a B2B campaign with different answer dynamics, and uses the vendor's 2-second threshold. Most enforcement benchmarks are calculated per campaign over a rolling 30-day period at roughly 3%, so the reported figure is not the number a regulator or plaintiff would compute. Rebuild the metric per campaign before drawing conclusions.
Once rebuilt, the root cause is structural. Eight of 60 agents are in training, so available capacity drops about 13% while the dialer keeps pacing at 2.1 lines per logged-in agent. The 12% human-answer rate makes the pacing algorithm over-dial to hit connect targets, which is why the spikes cluster in the training block. The four-calls-in-one-day complaint also points to attempt-cap enforcement failing across campaigns.
Next-shift changes:
- Cut pacing to 1.6-1.7 lines per agent until abandonment stays under 2.5% for five consecutive days.
- Route training blocks through agent-available or progressive mode.
- Enforce three attempts per seven days per consumer across all campaigns, not per campaign.
- Scrub against internal DNC and the complaint list immediately.
- Require abandoned-call playback with caller ID and callback number.
- Verify calling windows in each consumer's local time zone and log the complaint.
- Report abandonment daily per campaign, owned by the floor manager, escalated at 2.8%.
2. Build a 30-day monitoring and control plan. What metrics, thresholds, owners and evidence would you require?
Read the worked answer
A workable plan has three layers: daily operational control, weekly trend review, and a monthly audit that produces defensible evidence.
Daily, per campaign: abandonment rate (rolling 30-day and same-day), lines per available agent, agent availability, average speed of answer, attempt-cap breaches, and a count of complaints. Thresholds: same-day abandonment above 2.5% triggers an immediate pacing reduction by the dialer administrator; rolling 30-day above 3% triggers a written remediation plan owned by the operations manager and signed by compliance.
Weekly: a joint review of operations, workforce management and compliance covering staffing versus forecast, training-block scheduling, and a sample of 20 abandoned calls replayed to confirm the message and caller ID behave correctly. Also review DNC and revocation suppressions to confirm they propagated within 24 hours across every campaign.
Monthly: an independent audit that recomputes abandonment from raw dialer records using the campaign-level definition, checks that the 2-second answer threshold is configured as documented, and reconciles the internal DNC list against the dialer. Publish the audit to leadership.
Evidence to retain: raw dial logs, pacing-setting change history with timestamps and approver, recordings of abandoned-call messages, complaint log with resolution, attempt-cap reports, and DNC scrub certificates. Without change history you cannot show that a spike was corrected.
3. Sales insists on raising lines-per-agent to 2.6. Give your recommendation and the conditions under which you would approve a pilot.
Read the worked answer
Recommend against a floor-wide increase now. At 2.6 lines you would raise dial volume by roughly 24% while staffing capacity is temporarily down 13%. With a 12% human-answer rate, the marginal dials convert mostly into abandoned or very short answered calls, so the connect gain sales expects is unlikely and the compliance exposure is near-certain. Abandonment scales faster than connects when pacing is already above the staffing curve.
Approve a pilot only under these conditions:
- One team of at least 15 agents, fully staffed, no training block during the test.
- Two-week duration, with the preceding two weeks as baseline.
- Hard guardrails: campaign abandonment under 2.5% on every single day, no attempt-cap breaches, no repeat-call complaints. Any breach ends the pilot same day.
- Pre-agreed success metric: connect-to-conversation rate and sales per paid hour, not raw connects, so short junk calls do not count as wins.
- Compliance sign-off on settings before the first dial and a daily review during the test.
If the pilot shows a genuine lift in sales per paid hour while staying under guardrails, expand in 10-agent increments with a two-week observation window each time. The burden of proof sits with the pacing increase, not with the status quo.