Least Cost Routing That Doesn't Sacrifice the Call
Naive LCR picks the cheapest route and prays. Ours ranks every route per destination on a quality-weighted score (live ASR, ACD, NER, PDD, and CLI integrity, weighed against price), then fails over down the ranked list in milliseconds when a carrier blinks. An AI layer re-scores routes continuously, so the ranking you route on reflects the last minutes of real traffic, not last quarter's contract.
1. The problem with cheapest-first
Classic LCR sorts routes by price and takes the top of the list. It works until it doesn't: the cheapest route into a destination is frequently cheap because it's degraded: an oversold aggregator, a grey path losing CLI, a carrier silently dropping every third call. Your dialer sees ASR collapse, agents sit idle between connects, and the "savings" evaporate into wasted attempts and missed contacts.
The arithmetic is unforgiving: dropping ASR from 45% to 35% on a campaign costs far more in agent productivity than a $0.002/min rate difference saves. Routing decisions must price quality in, or they aren't cost decisions at all.
2. How DIDHub's LCR actually decides
Resolve the breakout
The dialed number maps to its precise breakout: not "Germany" but "Germany Mobile Vodafone." Every breakout has its own route portfolio and price points.
Score the portfolio
Each candidate route carries a live composite score: price, current ASR/NER, ACD trend, PDD percentile, CLI-delivery rate from probe verification, and capacity headroom. Scores decay fast; a route is what it did in the last minutes, not what its SLA claims.
Apply your routing profile
The same portfolio ranks differently per customer policy: quality-first (CLI-verified routes only, price secondary), balanced (the default), or cost-first (for traffic where presentation doesn't matter). Set per trunk or per destination.
Route, watch, cascade
The call goes to rank #1. Failure or timeout cascades to #2, then #3, inside the call path within milliseconds, invisible to your switch beyond a normal PDD.
3. The AI layer on top
Static scoring still reacts too slowly for wholesale voice, where a route can go from excellent to useless in minutes. The AI route-management system watches every route's live telemetry and intervenes without a human in the loop:
- Drift detection: ASR sliding 5 points over ten minutes on one breakout triggers de-ranking long before a threshold alarm would fire.
- Pattern recognition: FAS signatures, CLI-loss fingerprints, and capacity exhaustion each have distinct telemetry shapes the system has learned to catch early.
- Automatic traffic shifting: degraded routes drain gracefully to the next-ranked alternatives; recovered routes earn traffic back slowly, on probation.
- Carrier escalation: incidents open tickets with the underlying carrier automatically, with the telemetry attached.
Everything the AI does is visible to you: reroute events and the metrics that triggered them appear in your Analytics reports: observability we share, not a black box.
4. What this means for your operation
| Without quality-weighted LCR | With DIDHub |
|---|---|
| Campaign ASR collapses mid-run; you notice in tomorrow's report | AI shifted traffic minutes into the drift; the report shows the reroute, not the damage |
| You manage 8 carrier relationships and a failover spreadsheet | One trunk; the portfolio and cascade are our job |
| "Cheap" routes silently strip your CLI | CLI integrity is a scoring input, verified by probes |
| Quality data lives in the carrier's NOC | Per-destination KPIs in your dashboard, exportable |
Send us your traffic profile
Destinations, monthly minutes, CLI requirements. We'll return a route plan and rate deck within one business day.
FAQ
What is least cost routing (LCR)?
LCR is the routing logic that selects which carrier route each outbound call takes, traditionally by lowest price. Modern LCR, including DIDHub's, weighs live quality metrics (ASR, ACD, NER, PDD, CLI integrity) against price, because the cheapest route into a destination is frequently degraded.
Can I control the cost/quality trade-off?
Yes. Routing profiles (quality-first, balanced, cost-first) apply per trunk or per destination, so your brand-sensitive markets can run CLI-verified-only while notification traffic runs cost-optimized.
What happens when a route fails mid-call-attempt?
The call cascades to the next-ranked route automatically, inside the call path, within milliseconds. Your switch just sees a normal call setup.
How fast does routing react to a quality problem?
The AI layer works on live telemetry with decay measured in minutes: drift triggers de-ranking and traffic shifting typically before a human-monitored threshold alarm would have fired at all.
Is the routing visible to me or a black box?
Visible. Reroute events, the metrics that triggered them, and per-destination KPI trends are part of your Analytics reports.
Route your outbound through carrier-grade A–Z termination
Direct interconnects with some of the world's largest carriers, quality-weighted LCR, AI-managed route health, and the KPIs to prove it, shared with you rather than hidden.