Unified AI Workforce & Synthetic Customers: Cresta CMO's CX Vision

What happened

Cresta's Chief Marketing Officer has made a pointed intervention in the AI-versus-humans debate that continues to dominate contact centre discourse, arguing that the entire framing is a distraction. Speaking to CX Today, Cresta CMO Ping Banzon contended that the more productive question is not whether AI will replace agents, but how organisations can build what Cresta calls a unified AI workforce — one in which human and machine capability are deliberately orchestrated rather than positioned as rivals.

Banzon also surfaced a concept that is likely to generate significant discussion in CX and research circles: synthetic customers. Rather than waiting for real interactions to generate training data or surface service failures, Cresta is exploring the use of AI-generated customer personas to simulate conversations, stress-test agent and AI behaviour, and accelerate the feedback loop between product design and frontline performance. The implication is that the contact centre of the near future will be shaped not only by real customer demand but by modelled, synthetic demand running in parallel.

Why it matters

For CX leaders and service designers, the unified workforce framing carries a meaningful operational shift. If the majority of interactions already involve human–AI collaboration rather than one or the other acting alone, then the design challenge is no longer about adoption — it is about orchestration. That means rethinking workflows, handoff protocols, quality frameworks and performance metrics to reflect a blended reality rather than a binary one. Organisations still building business cases around "AI instead of headcount" are, on this reading, already behind the actual state of play.

The synthetic customer concept adds a behavioural economics dimension that deserves careful attention. Simulated personas can be tuned to represent specific psychological profiles — loss-averse callers, confused first-time users, high-churn-risk customers — allowing teams to rehearse emotionally complex scenarios without waiting for them to occur in the wild. Done well, this could compress the learning curve dramatically. Done poorly, it risks optimising for synthetic behaviour that diverges from how real people actually act under stress or uncertainty.

By the numbers

• 76% of contact centre interactions are currently handled by humans and AI working in tandem, according to Cresta's workforce report as cited by CMO Ping Banzon.

The Renascence take

The "replacement" narrative has always been more useful to vendors selling fear than to operators building sustainable service models. What Cresta is pointing toward — even if the language of "unified workforce" still carries a product-marketing flavour — reflects something more structurally important: that the unit of CX design is no longer the individual agent or the individual bot, but the system they form together.

Most readers will focus on the synthetic customer headline and miss the quieter implication: if 76% of interactions are already hybrid, then every quality score, CSAT survey and coaching programme built around purely human performance is measuring the wrong thing. The behavioral principle here is attribution error — we keep crediting or blaming the human when the outcome was co-produced. A customer-obsessed operator should immediately audit whether their performance frameworks account for AI contribution, and pilot synthetic customer simulations not as a novelty but as a structured rehearsal tool for high-stakes, emotionally charged interaction types — the ones where getting it wrong costs loyalty, not just a ticket.