Every January, we get flooded with predictions about AI growth skyrocketing and customers wanting better experiences. Thanks for the insight.
Here’s what I think is actually happening—and what will matter in 2026.
2025 was the year AI got serious
Behind all the “agentic AI” marketing buzz, real progress happened. We moved past the assumption that larger context windows and more parameters would solve everything. Models learned to think, not just respond. Tool use emerged—AI that can actually interact with other systems and update itself.
Jensen Huang made the distinction that’s now becoming obvious: consumers need generalized intelligence—assistants that can help with anything. Businesses need specialized intelligence—AI that deeply understands their specific domain, customers, and operations.
The real shift in 2026
Companies will finally mature past the “AI = automation = fewer humans” equation. That framing was always too narrow.
AI is a new technology that unlocks capabilities we’ve never had. That means creating roles that don’t exist yet. It means learning to approach problems from a non-deterministic perspective—which is uncomfortable for an industry built on scripts and rigid processes.
In contact centers specifically, this plays out in three ways:
1. We move beyond the broken metrics patchwork
For decades we’ve stitched together QA scores, CSAT, AHT, and service levels to approximate success. In reality, all of these metrics were influenced by human bias and organizational goals. I’ve seen customers selectively determine who to survey for CSAT based on the probability of a positive outcome.
QA had its time and place—it was the best we could do with the tools we had. But it’s been stretched way beyond what it was designed for.
AI-powered analytics tools like Spearfish change the game entirely. Instead of editing forms to collect new metrics on a go-forward basis, you can freely explore what’s actually happened with your customers—both historically and in real time. You can correlate related data from CRM, sales, and ticketing systems. And you’re reviewing 100% of interactions, not a small sample that may or may not represent reality.
2. Dispositions finally die
You know the drill—the ever-growing picklist where agents select one code that’s supposed to magically capture both customer need and interaction outcome.
I had a customer with over 1,000 items in their disposition list. I’ve monitored recordings where agents spent 4 minutes just selecting a code. And here’s the reality: agents choose the items they’re most familiar with, which usually comes directly from whatever class they were trained in. The data was never accurate.
AI can actually understand what happened in a conversation without forcing humans to categorize it badly. Intent detection replaces the checkbox.
3. Policies get a fresh look
The AI era—like the internet era before it—forces companies to revisit policies that only existed because of old technology constraints.
Here’s an example: instead of paying humans to screen every caller for fraud, what if we trusted our customers by default and used AI as the safeguard to catch actual bad actors? That has to be better than spending $5-10 per call on staffing and technology to treat everyone like a suspect.
The old model was: design a single rigid process for everybody because you couldn’t manage exceptions at scale. The new model is: trust more, verify smarter, handle exceptions better than ever.
Looking ahead
Once companies internalize these shifts, the real question becomes: what do we build that wasn’t possible before? Not “how do we do the same things cheaper”—but what entirely new capabilities can we create?
That’s the conversation worth having in 2026.
Ready to extract the hidden intelligence in your contact center? Schedule a contextual intelligence assessment and discover what your top performers already know.
About the Author: Ray Bohac is CEO and Co-Founder of Spearfish. Previously, he co-founded CallCopy/Uptivity (acquired by NICE Systems) and MotionCX, pioneering VoIP and speech analytics technologies. He’s spent 20+ years building and optimizing high volume contact centers for mid-market through Fortune 500 companies.
