A call center.
Photo: Kuni Takahashi/Bloomberg via Getty Images

Chatbots were a bust. But call center automation is real.

Protocol Enterprise

Welcome to Protocol | Enterprise, your comprehensive roundup of everything you need to know about the week in cloud and enterprise software. This Monday: the future of the call center, Amazon investigates AWS, and finance chiefs want a break from Excel.

(Was this email forwarded to you? Sign up here to get it in your inbox every week.)

The Big Story

Chatbots are dead, long live chatbots

For a brief moment in time, the marketing hype around chatbots made it seem like the future was at our fingertips.

No longer would we, as consumers, have to suffer through the agony of having to talk to another human. And all those call center agents who were no longer spending the day answering phones would, apparently, be suddenly freed up to do … something else.

There was just one problem: The tech didn't really work.

  • It's clear if you've called into a customer service center in the past few years that a robot can't handle all of your problems. And organizations still employ hundreds if not tens of thousands of agents who are still primarily tasked with answering calls.
  • So what happened? A simple case of the marketing hype overshadowing the maturity of the technology, according to industry experts.
  • "For the last several years, clients would believe anything that any vendor told them that AI was going to do for improving customer service," said Gartner analyst Drew Kraus. "A whole bunch have deployed chatbots thinking that they were going to get a virtual customer assistant … [and] what they ended up getting was expensive, glorified FAQs."

Despite that setback, there is a huge wave of digitization happening in the industry. And while the capabilities might not yet be at the level advertised by some vendors, most insiders believe the tech will soon catch up.

  • "The market has moved to a higher expectation," said Genesys CEO Tony Bates. It "isn't just the beginnings of automation in the call center, but technology that can allow you to orchestrate experiences that go beyond customer experience."
  • The big changes happening right now are perhaps a bit less exciting than robot agents. But they are necessary to reach the level of automation that providers are promising to clients.
  • For years, companies operated their call centers from on-premises data centers, tapping vendors like Genesys and Avaya that locked customers into multiyear contracts. That's quickly changing to the increasingly common, if not required, subscription model.
  • Many organizations, namely smaller ones with fewer than 150 agents, had already been moving those systems to the cloud. But it was clear when the pandemic hit and sent thousands of call center agents in large enterprises to their homes that the investment was important.

The market potential right now, like in so many other parts of enterprise tech, is huge. By the end of 2021, contact-center-as-as-service, or CCaaS, will account for just 32.3% of the total cloud center market, according to Gartner.

  • That's one reason why Zoom purchased Five9 for $14.7 billion. When it comes to large enterprise deals, analysts say the battle is between Five9, NICE inContact and Genesys. And all three are trying to achieve the same vision.
  • On top of automating aspects of the service process (i.e., chatbots) and giving tools to agents to do their jobs better, enterprises want to be more proactive in reaching out to those customers who are unhappy or dissatisfied, well before it gets to the point where they have to call a support agent.
  • They also want to be able to send targeted marketing offers to clients. And given the pivot away from third-party data, organizations see huge value in analyzing those calls for information that can help them build digital replicas of customers.
  • Today, that's largely not happening. For most people, when an agent answers the call, it has no information about who the customer is or their past interactions with the company. And much of the data is still segmented between departments.
  • "Sales teams and marketing teams and customer service teams … they still largely operate pretty separately," said Kraus. "There are some dynamic organizations that do a good job of sharing customer data, but there's a whole lot still that don't."

That market opportunity ahead is why companies like Genesys and Five9 are investing heavily in tools that are critical to that future. It's also why AWS is so interested in the space.

  • The company doesn't offer call center software in the traditional sense, which puts it in a different camp than Genesys and Five9.
  • Instead, AWS gives customers — who are often developers, not contact-center leaders — the option to build their own system using drag-and-drop modules and pre-built APIs.
  • "Amazon Connect has some very large deployments," said Kraus. But a lot of companies "aren't interested in that DevOps approach. They want that packaged application. They don't want their IT people involved in application development," he added.
  • On top of Connect, the company's AI tools, like Polly and Contact Lens, are widely used in the industry.

So for now, we'll all have to deal with the inevitable soul-crushing experience that is calling in to most service centers. But rest assured, that's changing.

  • And who knows, within the next few years customers may be yelling obscenities at a robot to fix their broken dishwasher or adjust their airline ticket instead of a human.
— Joe Williams


Outages aren't a matter of if, but when. According to data from ThousandEyes, global disruptions in March 2020 — when we saw remote work roll out at scale — were 63% higher than they were in January 2020.

Learn more

This Week On Protocol

Amazon digs into discrimination claims at AWS: After 550 employees raised issues with the cloud provider's current system to review complaints of systemic racism or bias, Amazon tapped an independent investigator to look into the allegations.

Five Questions For...

Erica Schultz, president of field operations, Confluent

Schultz was on our recent list of 10 people defining the new database landscape. Read the whole list here.

What was your first foray into the world of databases?

I started at Oracle as an SDR, straight out of college in the mid-'90s. It was the height of the database wars between Oracle, Sybase and Informix. I was a Spanish and Latin American Studies major in college, and I quickly realized I had to learn another new "language" in order to speak fluently about relational databases. The experience taught me that a winning product needs to be coupled with clear, differentiated, persona-specific messaging, as well as rigorous sales enablement and certification, in order to win big in the market.

What excites you the most about the future of the industry?

We are in the very early innings of a tectonic shift in the industry as enterprises accelerate their efforts to win in a digital-first world. To do this, they must leverage data as one of their most important assets, as well as connect all of the applications, systems and data layers as a real-time central nervous system. Organizations are beginning to realize that modernizing their data infrastructure — and setting their data in motion — is key to their next chapter of success.

What's your advice to younger technologists who want to build a career in this field?

Make your aspirations known and surround yourself with people who are supportive of those aspirations. When I was growing my career at Oracle, I felt a strong sense of belonging and support, as well as believed in the meritocracy that drove me to focus on performance. I've also come to believe that there are inherent systemic biases even in meritocratic systems, so self-advocacy and enlisting advocates is important.

What's the biggest hurdle companies are going to face in becoming a data-driven enterprise?

The biggest hurdles are on the people and culture side. For example, does an organization have the skill sets required to take advantage of new capabilities enabled by software? Can you upskill current talent and source new talent to meet this need? Organizationally, how do roles and responsibilities need to evolve given these new capabilities? What should the new operating model look like?

What's one piece of reading that you think should be a requirement for those in the industry?

"Growth Mindset" by Carol Dweck [Editor's note: While not a book title, Dweck writes often about the "growth mindset" concept] and "Grit" by Angela Duckworth.

Around the Enterprise


In hybrid work, the Internet is your new enterprise network. To understand what happened when a notable outage occurs, and what you can learn from it, subscribe to the ThousandEyes Internet Report, a weekly podcast that uncovers what's working, and what's breaking on the Internet—and why.

Learn more

Thanks for reading — see you Thursday!

Recent Issues