Practice · Contact Center & CX
Contact center and CCaaS selection led by people who have run the floor — a 2,000-user support operation at a 1.5% abandon rate, a call center designed and built, and an AI support agent tested against real volume.
Why This One Is Different
Every CCaaS platform demos well. What separates a good contact center from a bad one is routing design, queue structure, the metrics the floor is actually managed to, and whether the agents will use the tooling. None of that is in the vendor comparison, and all of it is decided after you sign.
Steve Staniszewski has run the floor. He took over a national helpdesk serving 2,000 users, rebuilt the team to 20, introduced incident priorities and categories, built the metrics and the executive reporting, and reworked the phone IVR for routing — landing a 1.5% call abandon rate and a 20% reduction in application incidents through root-cause analysis. He later helped design a call center and data center in El Paso and proved the VOIP platform in a lab before it was deployed.
David LaRue built a contact center at SIRVA, a $2.35B global moving and relocation group, and later owned customer experience across web, retail and marketplaces at KEH — where an AI support agent was prototyped against a live workload rather than a slide.
Engagements
| Organization | Category | Scope | Result |
|---|---|---|---|
| Empire TodaySteve · Helpdesk Manager · 2006–07 | Support Center · IVR | Took over the national technical and business-applications helpdesk serving 2,000 users. Rebuilt the team to 20, restructured the department, introduced incident priorities and categories, and built the metrics and executive reporting behind it. Reduced application incidents 20% through root-cause and trend analysis, and reworked the phone IVR for more efficient call routing. | 2,000 Users1.5% abandon rate |
| Empire TodaySteve · Telecom Manager · 2007–08 | Call Center Design · VOIP | Assisted in the design of the El Paso call center and data center, and built a Cisco IPT voice-over-IP test lab on Dell servers and VMware to prove the platform before deployment. | Site Designed |
| SIRVADirector, Global Head · 2004–08 | Contact Center Build | Built a contact center for a $2.35B global moving and relocation group, alongside global responsibility for business intelligence, web marketing, application development and mobile. | Built |
| AVSCSteve · Director, IT Ops · 2001–05 | Support Org | Reorganized the helpdesk alongside telecom, database and operations, within a 22-person department running on a $7M budget. | 22 Staff |
| Designed Alloy ProductsSteve · Director IT · 2008–12 | Helpdesk Build | Created and maintained the technical helpdesk as part of building the IT function from scratch across every location. | Built From Scratch |
| KEH CameraCTO / CDO, then CIO · 2020–24 | CX · AI Support | Owned customer experience across web, retail and marketplaces, and prototyped an AI customer-service agent against a live support workload. | AI Prototype |
What We Evaluate
Abandon rate, first-contact resolution and handle time are set by queue and routing design far more than by platform capability. We score how well a platform supports the routing you actually need, not its feature count.
Seat licenses, concurrent versus named, telephony carriage, recording, quality management, workforce management and the analytics module are usually priced separately. The headline per-seat figure rarely survives being assembled into a working configuration.
Deflection is the most oversold claim in this category. We score containment against escalation quality, and against what the system does when it is outside what it learned — the lesson from the service bay still applies.
A contact center is only as good as the customer, order and inventory data it can see. CRM, ERP and order management integration gets scored as a first-class requirement, because screen-popping the wrong record is worse than no screen pop.