Q&A with Michael Hutchison, Customer Operations Principal, eClerx
There’s a growing disconnect in customer experience right now that most companies still don’t fully understand. Customers move through the world with an expectation of continuity – an assumption that the business they’re interacting with should know who they are, what just happened, and what they’re trying to accomplish. But inside many organizations, the reality is still fragmented systems, siloed teams, disconnected workflows, and operational structures that were never really designed to support continuity in the first place.
What makes this moment interesting is that the technology is no longer the primary limitation. The tools to analyze interactions, connect journeys, surface patterns, and automate insight already exist. The harder problems now are operational and organizational. Companies have more visibility into customer behavior than ever before, but many still struggle to turn that insight into coordinated action.
That idea sits at the center of this conversation with Michael Hutchison, Customer Operations Principal at eClerx. What follows isn’t another discussion about AI magically transforming the contact center. It’s a grounded look at where the real friction still exists: the handoff points where context disappears, the interpretation gap between data and action, the hidden operational risks buried inside unreviewed interactions, and the growing realization that interaction intelligence is no longer just a QA function. It’s becoming an operational layer for customer experience itself.
And perhaps most importantly, this conversation surfaces a subtle but important shift underway inside customer operations: insight alone has very little value. The companies that pull ahead over the next few years won’t necessarily be the ones with the most data or even the most sophisticated AI. They’ll be the ones that can connect insight to action faster than everyone else.
1. Expectation Gap
Where is the biggest gap between what customers expect and what companies can actually deliver today?
The biggest gap is around consistency and continuity. Customers expect that every interaction – whether it’s digital, voice, or in-person – picks up where the last one left off. Most organizations aren’t there yet. Data is still fragmented, and systems don’t always talk to each other, so customers end up repeating themselves or starting over.
Is that gap more about technology, or about organizational alignment?
It’s both, but I’d lean toward organizational alignment. The technology exists to connect journeys, but teams are often structured in silos – digital, contact center, field operations – and they’re measured differently. That makes it harder to deliver a unified experience.
Where do you see that gap show up most visibly in the customer journey?
It shows up most clearly at handoff points – when a customer moves from self-service to an agent, or from one department to another. That’s where context gets lost and frustration builds quickly.
If a company could fix one thing tomorrow to close that gap, what would it be?
I’d focus on shared visibility into the customer journey. Even a simple step like ensuring agents can see what the customer just tried to do online makes a noticeable difference.
2. Signal vs Noise
Where do organizations get overwhelmed, and how should they prioritize signal over noise?
Organizations get overwhelmed when they move from limited sampling to analyzing everything. Suddenly there’s more data than anyone knows what to do with. The mistake is trying to act on all of it.
The way forward is to focus on repeatable patterns that impact outcomes – things like recurring failure points, compliance risks, or drivers of repeat contact.
What’s an example of data that looks useful but isn’t actually actionable?
Things like sentiment scores on their own. It’s helpful to know a call was negative, but unless you understand why it was negative and whether it’s happening consistently, it’s hard to act on.
What does ‘signal’ look like in a real operational setting?
Signal is when you can point to something specific – say, a billing issue driving 15% of repeat calls – and actually fix it. It’s tied to a root cause and a measurable outcome.
Who owns that interpretation layer – the technology or the business?
Technology can surface patterns, but the business has to interpret and act on them. That’s where experience and context come in.
3. QA to Revenue
At what point does interaction intelligence become a true revenue lever?
It becomes a revenue lever when it moves beyond scoring interactions and starts influencing behavior – how agents handle conversations, how issues are resolved, and how opportunities are identified in real time.
What’s the first metric that changes when this starts working?
Usually, you see improvement in first-contact resolution and a reduction in repeat calls. That’s often the first sign things are moving in the right direction.
Do you see more impact on retention, upsell, or cost avoidance?
Retention tends to move first. When experiences improve, customers are less likely to leave. Upsell comes later, once trust is established.
Can you point to a real example where this directly drove revenue?
We worked with a telecom client where better visibility into interactions helped identify patterns around billing confusion. Fixing that reduced churn and improved retention. It wasn’t framed as a revenue initiative initially, but that’s where the impact showed up.
4. Risk Exposure
If a company is still reviewing only 1–2% of interactions manually, what risks are they carrying?
They’re carrying a significant amount of unknown risk – compliance issues, poor customer experiences, and operational inefficiencies that simply aren’t being seen.
How would you quantify that risk in business terms?
It shows up in areas like churn, regulatory exposure, and repeat contact costs. If you’re only seeing a small fraction of interactions, you’re making decisions with incomplete information.
What’s the most common issue that goes undetected?
Consistency issues. One agent might handle something perfectly, while another creates friction – but without full visibility, it’s hard to spot those patterns.
Have you seen a situation where that blind spot caused real damage?
Yes – particularly in compliance-heavy environments. Issues that seemed isolated turned out to be widespread once analyzed at scale, and by then the impact had already built up.
5. Agent Impact
When AI is evaluating every interaction, how does that change the role of frontline agents? Does it improve performance or create pressure?
It does both, depending on how it’s implemented. When done well, it raises performance because agents have better visibility, clearer expectations, and more consistent feedback. When done poorly, it can feel like constant surveillance. The difference comes down to transparency and how insights are used – whether they’re there to support agents or simply measure them.
What new skills will matter most for agents going forward?
Judgment, adaptability, and communication. As routine work is automated, agents will spend more time handling complex or sensitive situations. The ability to interpret context, make decisions, and communicate clearly will matter more than following a script.
How should organizations rethink coaching in an AI-driven environment?
Coaching needs to become more continuous and more targeted. Instead of periodic reviews based on small samples, leaders can now focus on specific patterns and behaviors in near real time. The goal should be to guide improvement, not just evaluate performance.
6. The Next Capability
Over the next 24–36 months, what is the one capability every contact center will need that most don’t have today?
The ability to connect insights to action in real time. Most organizations can generate insights today, but very few can operationalize them quickly across teams and workflows.
What happens to companies that don’t build that capability?
They’ll fall behind on both cost and experience. Issues will take longer to resolve, and inefficiencies will compound over time.
Is this something they can buy, or do they have to build it?
It’s a combination. Technology can enable it, but it requires process alignment and operational discipline to make it work.
What should leaders be doing now to get ahead of it?
Focus on data quality, integration, and clear ownership of insights. Without those foundations, even the best tools won’t deliver meaningful impact.
The most interesting thing about this conversation isn’t the AI itself. It’s the emerging realization that customer operations is becoming an interpretation problem. Companies already have enormous amounts of customer data, interaction data, workflow data, and operational telemetry. The real differentiator now is whether an organization can turn that information into coordinated action before friction compounds into churn, cost, or customer distrust.
That shift has major implications. It changes the role of QA. It changes the role of frontline management. And it changes how companies think about operational design itself. The organizations that pull ahead over the next few years probably won’t be the ones with the flashiest AI demos. They’ll be the ones that can maintain continuity, preserve context, and operationalize insight faster than everyone else.
Photo by Gilles Lambert on Unsplash
