Why Omnichannel Customer Service Still Feels Disconnected
In this article
I still remember that beep.
If you worked in a call center years ago, especially on the receiving end of transferred calls, you probably know what I mean.
A call comes in. You answer, introduce yourself, and start the usual verification process.
"May I have your name, please?"
And before you can even finish, the customer is already shouting.
"I ALREADY GAVE THAT TO THE PREVIOUS AGENT!"
There you are. You have barely started your call, and you already have an angry customer.
You don't know what happened during the previous conversation. You don't know what the other representative promised or what troubleshooting has been done. You are expected to take over, but you are practically starting from scratch.
Meanwhile, the customer has probably been on the phone for quite some time. They have explained their concern, answered questions, and may have been put on hold before being transferred to you.
And now you are asking the same questions.
I experienced this more times than I care to remember during my years on the frontline.
It wasn't pleasant for the customer, and it certainly wasn't pleasant for the representative receiving the call.
I remember wishing that when a transferred call arrived, the previous representative's notes would automatically appear on my screen.
Something that would tell me the customer had already been verified, what the concern was, and what had been done about it.
Instead of starting another round of questions, I could acknowledge the concern and keep helping.
Of course, there are legitimate reasons to repeat certain security checks. But there is a difference between necessary verification and asking customers to repeat information simply because our systems cannot communicate.
Back then, having all that information appear automatically felt almost like science fiction.
Today, much of that technology is readily available.
And yet customers are still being asked to explain the same problem over and over again.
So what exactly have we improved?
More Support Channels, But Are We Actually Connected?
Businesses have come a long way in customer service.
We used to rely heavily on telephone support. Today, customers can reach companies through email, live chat, social media, mobile applications, messaging platforms, and AI-powered chatbots.
The idea behind omnichannel customer service is reasonable. Customers should be able to choose how they want to communicate and move between channels without losing the continuity of their conversation.
But offering multiple channels is not the same as connecting them.
A customer might start with a chatbot, move to live chat, and eventually request a human representative.
If that representative has no idea what happened during the previous interactions, have we really improved the experience?
We have given the customer more options to contact us, but perhaps not a better way to resolve the problem.
This is not just my observation.
The 2026 State of Customer Experience Report by Genesys, which surveyed 5,811 consumers and 1,560 CX and business leaders worldwide, found that 95% of consumers consider it important that their information follows them when switching between service channels.
Yet 48% of surveyed organizations reported that information previously collected was not automatically passed between virtual and human agents.
Think about that.
We have systems sophisticated enough to understand customer questions, automate responses, and route inquiries. But many organizations still struggle to ensure the next representative knows what the customer has already said.
And I believe this problem deserves more attention because it affects more than customer convenience.
It affects the people working on those interactions, too.
An agent receiving a frustrated customer without any context must spend time recovering the conversation before attempting to resolve the issue.
That takes time.
And as anyone who has worked in a contact center knows, we are constantly measured against time.
Six Functions. One Representative. Six Minutes and Thirty Seconds.
Another experience from my earlier BPO career comes to mind.
During my years with Convergys, handling the Sprint account, we had an operating arrangement involving six different functions or splits.
As far as I can remember, these covered Customer Service, Billing, Welcome Setup, Technical Support, Collections, and probably accounts tagged for fraud-related concerns.
I cannot even remember the last one with complete certainty. That is how long ago it was.
What I do remember clearly was the expectation.
One representative handled those six functions.
We were not supposed to keep transferring customers between departments just because their concerns involved different areas of the business. Unless the situation required an escalation, such as supervisor intervention, the representative was expected to handle the concern.
From the customer's perspective, that actually made sense.
If I call about a technical issue and discover that it also affects my bill, why should I have to speak with another representative?
Wouldn't it be easier if the person already assisting me could handle both?
That was one of the strengths of the arrangement.
Now comes the interesting part.
Our Average Handle Time target was approximately six minutes and thirty seconds.
Yes. Six minutes and thirty seconds.
You were expected to handle concerns across six different functions, navigate the necessary systems, understand the problem, complete the required actions, explain the resolution, and document the interaction.
Not all six functions on every call, of course. But you needed to be prepared for whichever concern came your way, sometimes involving more than one function.
And you still had that AHT target.
Then came Customer Satisfaction .
You were expected to deliver excellent CSAT scores by the end of the week.
Good luck with that.
Imagine receiving a customer with a complicated technical concern that also involves a billing dispute.
You investigate the technical issue, explain the billing implications, make whatever corrections you are authorized to make, and ensure the customer understands what happened.
You want to help. You also want to resolve the concern without another transfer.
But somewhere in the back of your mind, you know that every additional minute affects your AHT.
And if you rush the interaction, your CSAT could suffer.
Looking back, I can appreciate what that arrangement was trying to accomplish. It reduced unnecessary transfers and gave representatives greater responsibility for resolving customer problems.
But it also placed considerable demands on the person answering the call.
This was a historical experience from my time on the account, not a description of how those companies operate today.
Still, it taught me something that I think remains relevant.
Removing the need to transfer customers does not automatically eliminate operational complexity. Sometimes, we move that complexity from the customer to the representative.
Unless we account for that additional responsibility, we may create another problem while trying to solve the first one.
When Good KPIs Do Not Necessarily Mean Good Customer Experience
I have spent enough time in training, quality assurance, and operations to understand why contact centers need performance metrics.
Average Handle Time, Service Level, First Contact Resolution, Quality Assurance, and Customer Satisfaction are all important.
We need these numbers to manage staffing, maintain service standards, identify problems, and control operational costs.
I am not suggesting that we stop measuring them.
What I question is whether we always interpret them correctly.
Take AHT.
Suppose a representative handles a call in four minutes and transfers the customer to another department. The second representative spends another five minutes handling the concern.
Depending on how performance is measured, both representatives might meet their individual AHT targets.
Now compare that with a cross-trained representative who takes eight minutes to resolve the same concern without transferring the customer.
Which one was more efficient?
It depends on the actual work performed, whether the issue was resolved, the resources consumed, and what happened afterward.
The second approach may have required fewer interactions and less effort from the customer. But it would be a mistake to assume that automatically makes it cheaper or more effective in every situation.
We need to examine the complete journey.
My experience with Sprint showed me one side of the problem. Expecting a representative to handle multiple functions can make the experience easier for customers, but it also makes the representative's job more demanding.
My experiences receiving blind transfers showed me the other side. Dividing work between teams can simplify individual responsibilities, but poor coordination can leave customers repeating themselves and agents managing frustration they did not create.
Neither arrangement is automatically right or wrong.
The problem arises when we measure efficiency without fully considering the consequences of how we organize work.
This becomes even more important when different departments or service providers are involved.
The first department may have met its service level. The second may have completed the required verification. The third may have finally resolved the concern.
Each department can present a report showing it completed its responsibilities.
But the customer may have spent an hour trying to resolve something that should have taken much less time.
And somewhere in those reports, that experience can disappear.
The concern about disconnected customer experiences also appears in Five9's 2026 Business Leaders Customer Experience Report.
The research found that nearly all surveyed CX decision-makers believed their organizations effectively preserved context during AI-to-human handoffs.
Yet 83% of surveyed consumers said they had to repeat information at least sometimes after being transferred.
These were separate respondent groups, not a direct comparison of the same customer interactions. Nevertheless, the difference between management perceptions and consumer experiences is worth examining.
Perhaps the question is not simply whether our teams are meeting their KPIs.
Perhaps we should also be asking whether those KPIs reflect what customers and frontline representatives actually experience.
We Design Operations Around Functions. Customers Don't Think That Way.
Companies divide customer service work into different functions for good reasons.
Billing requires particular knowledge. Technical Support requires troubleshooting skills. Collections involves different procedures. Fraud-related cases may require specialized access, investigation, and security protocols.
Specialization is not necessarily a bad thing.
Neither is cross-training representatives to handle several types of concerns.
Both approaches can work.
What matters is whether the operating model supports the experience the company wants to deliver.
If we divide work between teams, we need a reliable way to preserve information and progress when a concern moves between them.
If we assign several functions to one representative, that person must receive the training, tools, authority, and time needed to do the work properly.
Otherwise, we are simply choosing where the inconvenience will fall.
This is especially relevant in outsourcing.
A BPO provider might be contracted to manage customer service while the client retains control of billing approvals, technical escalations, account adjustments, or other back-office functions.
The provider can meet its contractual KPIs and still have customers experiencing difficulties because certain information or decisions are controlled elsewhere.
On the other hand, a provider cannot automatically blame the client's systems for every service failure.
Representatives must still be properly trained. Cases must be documented accurately. Escalations must be handled correctly.
Both parties have responsibilities.
The client and the provider should be working toward the same customer outcome, rather than simply defending their respective performance reports.
After all, the customer does not really care whether the person answering the call works directly for the company or for its outsourcing partner.
They called one business expecting help.
That is the experience we need to design around.
A Better Customer Journey Should Make the Agent's Job Easier, Too.
I believe customer journey mapping deserves a more practical approach than simply drawing the stages a customer goes through.
We need to examine what actually happens at each stage.
What information does the customer provide?
Who receives it?
Where does it go?
If the concern is transferred, can the receiving representative see what has already been done?
If the same representative is expected to handle several functions, do they have everything they need to do that efficiently?
This is where I think organizations sometimes miss an opportunity.
We talk about reducing customer effort, which is important. But we should also be looking at the unnecessary effort we place on our employees.
Think about that blind transfer I described earlier.
If the receiving agent had access to the previous representative's notes, the conversation could begin at a completely different point.
Instead of spending several minutes asking questions the customer has already answered, the agent could focus on resolving the concern.
The customer would be less frustrated.
The representative would not have to spend the beginning of the call calming someone down over a problem they did not create.
The time saved could also improve operational performance.
Now consider the Sprint example.
If a representative is expected to handle several different functions, the organization should ensure the systems are easy to navigate, the processes are clear, and the AHT expectations match the complexity of the work.
In both cases, better customer journey design can reduce unnecessary effort.
Not by removing legitimate procedures or lowering service standards, but by ensuring that the work makes sense for the people involved.
That is the part of customer experience I believe deserves more management attention.
A customer journey is not well designed if it becomes easier for the customer only because the representative is expected to absorb every difficulty behind the scenes.
It needs to work for both.
What Should CX Leaders Be Looking At?
I would not recommend throwing away the existing contact center scorecard.
Instead, I would review it alongside a few additional operational questions.
First, how much unnecessary effort are we asking customers to make?
If customers repeatedly provide the same information, contact several departments, or follow up on unresolved concerns, those are warning signs that the customer journey needs attention.
Second, what does the representative actually have to do to complete the work?
We cannot fairly evaluate a simple account inquiry and a complicated technical dispute as if they require the same effort.
Likewise, an agent expected to handle multiple functions may need a different training approach, level of support, and productivity benchmark from someone handling a narrower range of concerns.
Third, are we measuring the time required to finish an interaction or the effort required to resolve the customer's problem?
Those are not always the same thing.
A four-minute interaction that generates two additional calls may consume more organizational resources than a longer interaction that resolves the concern correctly.
That is why you should consider AHT alongside First Contact Resolution, repeat contact rates, total resolution time, quality, and customer satisfaction.
And I would add one more practical exercise.
Have someone from management follow a sample of actual customer cases from beginning to end.
Not just one recorded call.
Follow the entire issue.
See how many people handled it, which systems they used, where delays occurred, and whether any work had to be repeated.
Then ask the representatives involved what made the job unnecessarily difficult.
I suspect we would learn things that don't always show up on our dashboards.
The Technology Is Here. The Question Is How We Use It.
Looking back at my early frontline experiences, it is interesting to see how much customer service technology has evolved.
What I once wished would automatically appear on my screen is now possible through modern contact center platforms.
Customer interaction histories can be retrieved.
Conversation summaries can be generated.
AI can help identify the reason for a contact and provide representatives with information from earlier interactions.
Customer relationship management systems can connect records across different channels, provided the organization has implemented the necessary integrations and controls.
These capabilities can make a meaningful difference.
But buying the technology is only part of the solution.
A conversation summary is not particularly useful if nobody reads it.
A connected system cannot solve a problem if the representative lacks the authority to act.
And AI will not make an unrealistic AHT target reasonable simply because it can produce information faster.
We still need to design the processes properly.
We need to train employees, define responsibilities, establish appropriate security controls, and make sure performance expectations reflect the work that needs to be done.
Technology should make customer service easier to deliver, not merely give management more sophisticated ways to monitor it.
The Customer Sees One Company. The Representative Feels the Pressure.
I have had the opportunity to experience the BPO industry from different positions.
I started on the frontline. I moved through training, quality assurance, and management, and eventually into sales and business leadership.
Those experiences have given me a greater appreciation of why companies operate the way they do.
I understand the commercial pressure to control costs.
I understand why AHT matters and why companies invest in improving productivity.
I also understand the importance of CSAT, service levels, and meeting client expectations.
But I have not forgotten what it feels like to be the person answering the call.
To receive an angry customer who has already explained the same concern several times.
Or to be expected to handle a complicated issue across several functions while keeping one eye on the AHT target.
Those situations were not always the result of poor employees or unreasonable customers.
Sometimes, they were the result of how the operation was designed.
That is why I believe customer experience and employee experience cannot be treated as entirely separate conversations.
We cannot keep adding complexity to the representative's job and assume that it will not affect service delivery.
Nor can we continue expanding customer support channels without ensuring that information, progress, and responsibility move properly between them.
Customer concerns will always be complicated. Some will require specialized expertise, additional verification, or more time than we would prefer.
That is normal.
What should not be normal is making customers start over because our systems are disconnected, or expecting agents to compensate for poor processes while maintaining performance targets that fail to account for the work involved.
A truly connected customer experience should make the journey better for everyone involved.
The customer should not have to explain the same problem five times.
And the representative should not have to fight the operating model to deliver good service.
So the next time we review our contact center dashboard and see excellent numbers, perhaps we should ask two more questions.
How much unnecessary effort did our customers go through before their problems were resolved?
And how much unnecessary pressure did we put on our representatives along the way?
Because sometimes, the problem is not the person answering the call.
It is everything we have designed around that person.
