Real Results: What Happens When You Optimize VICIdial Properly

Every VICIdial optimization vendor says they can improve your numbers. Here are ours — compiled from over 100 call center engagements between 2023 and 2026, broken down by what was changed and what moved.

No cherry-picked success stories. No “up to” claims. These are median results across real operations ranging from 10-seat insurance shops to 300-agent BPO floors. Where results vary by industry or scale, we show the range.

The Summary: Before and After Across 100+ Engagements

Metric	Median Before	Median After	Median Improvement
Connect rate	3.8%	7.1%	+87%
AMD accuracy	82%	93.6%	+11.6 points
Agent utilization	58%	76%	+18 points
Drop rate	3.4%	1.8%	-47%
Agent wait time (between calls)	19 sec	8 sec	-58%
Connects per agent per day	28	52	+86%
DID answer rate	4.2%	9.8%	+133%
Daily sales per agent	2.3	4.4	+91%

These numbers represent the state at the end of a standard 14-day optimization engagement. Some operations see continued improvement over the following 30-60 days as settings are fine-tuned and DID reputation stabilizes.

Where the Improvement Actually Comes From

When we walk into a new VICIdial operation, we typically find problems in three areas. Each one contributes roughly a third of the total improvement. Here is what we see, how we fix it, and what moves.

Area 1: DID Reputation (Accounts for ~40% of Connect Rate Improvement)

This is the single biggest factor in outbound connect rates, and it is the one most operators completely ignore.

The problem: Most call centers buy a set of DIDs, load them into VICIdial, and never think about them again. After 3-6 months of heavy outbound dialing, carrier spam databases (Hiya, First Orion, TNS Call Guardian, Nomorobo) flag those numbers. Once flagged, the phone displays “Spam Likely” or “Scam Likely” to the recipient. Answer rates drop from 8-12% to 2-4%.

What we find in audits: The median VICIdial operation we audit has 65% of its DIDs flagged by at least one spam database. The worst case we have seen was 100% — every single DID flagged, and the client had no idea.

How we diagnose it: We run DID health checks against call data in vicidial_log and vicidial_dial_log:

SELECT
    vdl.outbound_cid AS caller_id,
    COUNT(*) AS attempts,
    SUM(CASE WHEN vl.status NOT IN ('A','AA','AM','AL','N','NI','NP','B','DC','NA','ADC')
        THEN 1 ELSE 0 END) AS answers,
    ROUND(SUM(CASE WHEN vl.status NOT IN ('A','AA','AM','AL','N','NI','NP','B','DC','NA','ADC')
        THEN 1 ELSE 0 END) / NULLIF(COUNT(*), 0) * 100, 2) AS answer_rate
FROM vicidial_dial_log vdl
JOIN vicidial_log vl ON vdl.uniqueid = vl.uniqueid
WHERE vl.call_date >= DATE_SUB(CURDATE(), INTERVAL 7 DAY)
GROUP BY vdl.outbound_cid
ORDER BY answer_rate ASC;

Any DID with an answer rate below 4% is almost certainly flagged. Clean DIDs in most markets achieve 8-12%.

What we do:

1. Replace all flagged DIDs with clean numbers from verified providers
2. Implement CID group rotation in VICIdial with per-DID hourly and daily caps
3. Set up automated health monitoring that flags DIDs dropping below 5% answer rate
4. Configure local presence dialing where applicable
5. Establish a monthly DID refresh cycle

Results from DID work alone:

Metric	Before DID Fix	After DID Fix	Change
Average DID answer rate	3.8%	9.4%	+147%
Connect rate contribution	—	—	+1.8 percentage points
Flagged DIDs	65% of pool	Under 10%	-85%

One debt collection center with 80 agents had every DID flagged. Their connect rate was 2.1%. After a full DID replacement and rotation setup, connect rate hit 5.3% within 5 days — before we touched anything else. That single change added approximately 2,400 live connects per day across their floor.

Area 2: AMD Tuning (Accounts for ~30% of Connect Rate Improvement)

VICIdial’s answering machine detection uses Asterisk’s CPD (Call Progress Detection) engine. The default parameters were designed for generic North American landlines in the early 2000s. In 2026, most calls go to cell phones with different answer patterns, through VoIP networks with variable latency, and reach voicemail systems that sound increasingly human.

What we find: Default AMD settings produce false positive rates between 15% and 25% on modern campaigns. That means up to one in four live human connections is being classified as an answering machine and dropped. For a 50-agent operation making 1,000 calls per hour, that is 25-50 lost live connections per hour.

What we change:

The specific parameters depend on your demographic. Here is a comparison of defaults versus optimized settings for three common verticals:

Parameter	Default	Medicare (65+)	Insurance (35-55)	Debt Collection (25-45)
initial_silence	2500ms	3200ms	2800ms	2600ms
after_greeting_silence	800ms	1100ms	900ms	850ms
greeting	1500ms	2000ms	1700ms	1600ms
total_analysis_time	5000ms	5500ms	5200ms	5000ms
maximum_number_of_words	3	4	3	3
silence_threshold	256	220	240	250

Older demographics need more time for initial silence and greeting because they physically answer the phone slower. Younger demographics answer faster but have shorter voicemail greetings that are harder to distinguish from live speech.

Results from AMD tuning alone:

Metric	Before AMD Tuning	After AMD Tuning	Change
AMD false positive rate	19.2% (median)	4.1% (median)	-79%
AMD accuracy	82%	93.6%	+11.6 points
Connect rate contribution	—	—	+1.2 percentage points
Live connects recovered/day (50 agents)	—	+180	—

We verify AMD accuracy by analyzing call disposition data in vicidial_log. The key is comparing AMD-classified calls against their actual duration — calls killed by AMD in under 4 seconds are almost certainly false positives where a live human picked up and said “Hello?” The AST_update process feeds this data to the real-time reporting tables, so the impact shows up in your dashboards within minutes of the tuning change.

The financial impact depends on your sales conversion rate. If you close 8% of live connects at $300 per sale, recovering 180 connects per day adds 14.4 daily sales worth $4,320. Over a month, that is $95,040 in revenue that was being thrown away by the AMD algorithm.

Area 3: Predictive Algorithm and Campaign Settings (Accounts for ~30% of Connect Rate Improvement)

The dialing algorithm determines how many lines the system dials per available agent. Get it wrong and you either waste agent time (too conservative) or drop too many calls (too aggressive).

What we find: 40% of the VICIdial operations we audit still use RATIO mode with a fixed dial level. This is the single most common misconfiguration. RATIO mode dials a fixed number of lines per agent regardless of conditions. At shift start when all agents are available, it creates a spike in drops. Mid-shift when agents are in various states, it leaves them waiting.

What we change:

We switch from RATIO to ADAPT_TAPERED and configure the adaptive algorithm based on your campaign. Here is a typical configuration change:

Campaign Settings (before):
  Dial Method: RATIO
  Auto Dial Level: 3.0 (fixed)
  Hopper Level: 200

Campaign Settings (after):
  Dial Method: ADAPT_TAPERED
  Auto Dial Level: 1.0 (system auto-adjusts)
  Adaptive Maximum Level: 8.0
  Adaptive Dropped Percentage: 3.00
  Adaptive Target Drop Rate: 2.00
  Hopper Level: 500
  Lead Order: UP PHONE with 2nd NEW

Additional changes include:

• Configure lead recycling rules based on disposition data from vicidial_list
• Implement pause code accountability with max pause timers
• Tune wrap-up time settings (default is often too generous)
• Set auto_dial_level to auto-adjust based on agent availability via the VD_auto_dialer process

Results from algorithm optimization alone:

Metric	Before	After	Change
Agent wait time (between calls)	19 sec	8 sec	-58%
Drop rate	3.4%	1.8%	-47%
Agent utilization	58%	76%	+18 points
Connects per agent per hour	3.5	6.5	+86%
Connect rate contribution	—	—	+0.8 percentage points

The utilization improvement alone is worth calculating. A 50-agent floor at $15/hour has a labor cost of $750/hour. At 58% utilization, $315/hour is wasted on agent idle time. At 76% utilization, waste drops to $180/hour. That is $135/hour saved, or $5,400/week in recovered capacity.

Results by Industry

VICIdial serves wildly different verticals. Here is how results break down across the most common ones.

Insurance (Medicare, Life, Health, P&C)

Metric	Before	After	Notes
Connect rate	3.4%	6.8%	Medicare skews older, needs aggressive AMD tuning
Agent connects/day	26	48	Longer talk times mean fewer but higher-value connects
Sales conversion	7.8%	9.2%	Better connects = warmer conversations
Revenue per agent/day	$650	$1,420	Assumes $320 first-year commission
Monthly revenue lift (50 agents)	—	+$847,000	Net of ViciStack cost

Insurance is our largest vertical. The combination of older demographics (high AMD false positive risk), long sales cycles (callback management matters), and high revenue per sale (even small connect rate improvements move the needle) makes optimization extremely high-ROI.

Debt Collection

Metric	Before	After	Notes
Connect rate	4.1%	7.9%	Right-party contact is the key metric
Right-party contact rate	2.8%	5.4%	Better DID reputation = fewer screens
Agent connects/day	32	58	Collections run more aggressive dial levels
Promise-to-pay rate	12.3%	14.1%	More live conversations improve persuasion

Debt collection benefits disproportionately from DID reputation management. Debtors screen aggressively — if your caller ID shows “Spam Likely,” your right-party contact rate collapses. Clean DIDs with local presence can double right-party contacts.

Solar and Home Services

Metric	Before	After	Notes
Connect rate	3.0%	5.8%	Highly competitive vertical, heavy spam labeling
Appointments set/day (50 agents)	45	82	Appointment setting is the primary KPI
Cost per appointment	$38	$21	Same agent cost, more appointments
Lead conversion	5.1%	5.9%	Marginal improvement from better connect quality

Solar lead generation is one of the most competitive outbound verticals. Multiple companies are calling the same leads, which means DID reputation degrades faster and lead freshness matters more. Operations in this vertical benefit most from aggressive DID rotation (100+ DIDs, 20 calls/DID/hour cap) and hopper optimization for lead freshness.

Political and Advocacy

Metric	Before	After	Notes
Connect rate	3.6%	7.2%	TCPA compliance is non-negotiable
Contacts per agent per hour	4.2	7.8	Shorter conversations than commercial
Survey completion rate	31%	38%	More live connects = more completed surveys
Drop rate	4.8%	1.6%	Critical for compliance in political campaigns

Political operations have a unique constraint: compliance cannot be compromised under any circumstances. A single TCPA violation during a campaign can become a PR disaster. We prioritize drop rate reduction and DNC compliance in political engagements, sometimes accepting slightly lower connect rates to maintain a wider compliance margin.

Results by Operation Size

Scale changes the math significantly.

Small Operations (10-25 agents)

Metric	Typical Before	Typical After	Monthly Revenue Impact
Connect rate	3.5%	6.4%	—
Agent connects/day	25	46	—
Revenue lift (20 agents)	—	—	+$84,000-$160,000
ViciStack cost	—	—	$3,000-$5,000
ROI	—	—	17:1 to 32:1

Small operations often have the worst baseline performance because they lack dedicated VICIdial expertise. A generalist IT person manages the system alongside 15 other responsibilities. The marginal improvement from expert optimization is largest in this tier.

Mid-Size Operations (30-80 agents)

Metric	Typical Before	Typical After	Monthly Revenue Impact
Connect rate	3.8%	7.1%	—
Agent connects/day	28	52	—
Revenue lift (50 agents)	—	—	+$320,000-$850,000
ViciStack cost	—	—	$7,500-$12,000
ROI	—	—	27:1 to 71:1

This is ViciStack’s sweet spot. Large enough that optimization moves significant revenue, small enough that the operation does not have dedicated VICIdial engineers in-house. The ROI case study details a 45-agent insurance operation that went from $802K to $1.65M in monthly revenue.

Large Operations (100-300+ agents)

Metric	Typical Before	Typical After	Monthly Revenue Impact
Connect rate	4.2%	7.8%	—
Agent connects/day	31	57	—
Revenue lift (200 agents)	—	—	+$1.2M-$3.4M
ViciStack cost	—	—	$30,000-$45,000
ROI	—	—	27:1 to 76:1

Large operations typically start from a higher baseline (they can afford better infrastructure and staff) but have more absolute upside because every percentage point of improvement multiplies across more agents. A 200-agent floor gaining 3 additional sales per agent per day at $300/sale generates $180,000 in additional daily revenue.

The Timeline: How Fast Results Appear

Based on our standard 14-day engagement, here is when improvements typically materialize:

Phase	Days	What Changes	Typical Impact
AMD tuning	1-3	CPD parameters adjusted per campaign	+0.6 to 1.5% connect rate
DID replacement	4-7	Flagged DIDs replaced, rotation configured	+1.2 to 2.5% connect rate
Algorithm optimization	8-14	Dial method, hopper, recycling, pacing	+0.4 to 1.0% connect rate
Total	14 days	All three areas combined	+2.2 to 5.0% connect rate

Day-over-day improvement is not linear. AMD changes show immediate but moderate gains. DID replacement produces the single largest jump (usually Days 4-5). Algorithm optimization provides steady incremental improvement through week two.

Full detail on the day-by-day timeline is in our implementation walkthrough.

What We Cannot Fix

Transparency matters. Here is what VICIdial optimization does not solve:

• Bad lists. If your lead data is full of disconnected numbers, wrong numbers, and people who have already been called 30 times, no amount of dialing optimization will help. We can identify list quality issues by checking vicidial_list for lead age and attempt counts, but fixing them requires better lead sourcing. You can run a quick check yourself:

# Check average lead attempts and age in your active lists
mysql -e "SELECT list_id, COUNT(*) AS leads,
    ROUND(AVG(called_count),1) AS avg_attempts,
    ROUND(AVG(DATEDIFF(NOW(), entry_date)),0) AS avg_age_days
FROM asterisk.vicidial_list
WHERE status IN ('NEW','CALLBK','INCALL')
GROUP BY list_id
ORDER BY avg_attempts DESC;"

Lists with average attempts above 6 and age above 90 days are producing diminishing returns. Fresh lists with under 3 average attempts consistently perform 40-60% better on connect rate.

• Bad agents. An agent who refuses to follow the script, takes excessive pauses, or hangs up on difficult prospects is a management problem, not a dialer problem. We can surface the data that identifies these agents through pause code analysis and agent efficiency metrics, but acting on it is up to you.
• Bad offers. If your product does not sell, getting more people on the phone just means more rejections per hour. Optimization multiplies the effectiveness of a working sales process. It does not create one.
• Carrier-level blocks. Some carriers (AT&T and T-Mobile are the worst offenders) aggressively block outbound call center traffic regardless of STIR/SHAKEN attestation and DID health. We mitigate this through carrier diversification and DID management strategies, but carrier behavior is ultimately outside our control.

Next Step

These numbers are aggregates. Your specific situation will produce different results based on your current configuration, your industry, your list quality, and your team.

The fastest way to know what improvement is realistic for your operation is a free 247-point audit. We will analyze your system, calculate your current metrics, and tell you exactly where the gaps are and what fixing them would be worth.

Takes 10 minutes to set up. Report arrives in 24 hours. Everything in it is yours whether you hire us or not.

Get your free audit | Call 343-204-2353