The Gist
NiCE landed its largest CXone and Cognigy deal to date. An eight-figure win with the British tax authority, delivered alongside Capgemini. It also beat its own guidance. Revenue and earnings both came in above the high end of the range NiCE set for itself. The stock fell anyway. Investors pushed back hard on questions about renewal discounts, decelerating AI growth, and how much of the “AI story” is actually NiCE’s own technology.
NiCE landed its largest CXone and Cognigy deal to date — an eight-figure deal with the British government. It also beat its own guidance.
That’s part of the good news from its second-quarter earnings report, delivered Aug. 5.
The challenging news: none of it moved the stock in NiCE’s favor. Shares fell roughly 6% the morning the results came out, and it was trending around 4% down by afternoon on the East Coast of the US. It’s the second straight quarter that’s happened. Q1 saw a stock drop despite strong numbers, too.
That’s the big question after the Q2 2026 earnings report and conference call. With AI investments growing, and adoption climbing, when will Wall Street’s investments catch up?
Let’s break down the questions analysts are asking of NiCE and how NiCE responds.
What’s Inside the Numbers for NiCE in Q2, 2026
First, the numbers:
Total revenue: $782.3 million. Up 8% year over year, above the high end of guidance. AI annual recurring revenue: $362 million, up 52% year over year, now 15% of cloud revenue. That growth rate is down from 66% just one quarter earlier. CFO Beth Gaspich confirmed Cognigy alone added roughly 250 basis points to cloud growth this quarter, meaning a real chunk of the “AI story” is the acquisition, not organic build.Cloud growth itself slowed to 12.6%, down from 14.6% in Q1. NiCE says the slowdown was intentional — a small number of discounted renewals traded for long-term AI commitments from marquee customers, not a demand problem, and it’s confident in its cloud backlog growth, which doesn’t include the big deal with HMRC of the British government.Customer wins. TripAdvisor with NiCE Cognigy AI agents went “from concept to its first live automated voice calls in just two and a half months,” Russell said, with its AI agent posting a 90% customer sentiment score against 71% for human agents. GX Bank in Malaysia now runs 95% customer satisfaction and 95% first contact resolution on the platform, with AI autonomously resolving 70% of chat interactions. What Analysts Asked — and What NiCE Answered
Some questions on the call get at why Wall Street keeps discounting NiCE’s beats. Each one pushes on a different piece of the “trust the AI story” case.
Is the AI Growth Number Cognigy, or Is It NiCE?
The sharpest exchange of the call came from an analyst standing in for Piper Sandler’s James Fish, who did his own math: “ARR is still growing over 50% but if we back out Cognigy here, it seems that that AI grew sub-30% compared to 40% last quarter. What went wrong with that organic AI offering slowing down?
Gaspich’s answer didn’t engage with the specific comparison: “You can’t really look at organic versus inorganic with respect to Cognigy. At the time we made the acquisition of Cognigy, it meant that we discontinued selling the other comparable solutions we had previously. So you must look at it in consolidation. It doesn’t really pertain to look at it in a segmented way at this point.”
Who Does NiCE Actually Compete on Major CX Deals?
NiCE was asked the question every buyer comparing vendors wants answered: on the large, eight-figure deals, who’s actually in the room? Genesys? Sierra and Decagon? Salesforce’s Agentforce Contact Center?
“The answer is probably all of the above,” NiCE CEO Scott Russell said on the earnings call today. “I mean, the reality is it is a competitive environment, and let’s face it, we fully acknowledge that all customers have choice of approach. We believe firmly that the system of engagement and the customer engagement platform, in a united manner, will be the preeminent way large enterprises and medium for that will be able to deliver their customer experience.”
Brands want AI-native solutions, he added, but then have to figure out how to integrate it and how to maintain that integration. Who’s got the end-to-end capability that provides voice, digital, human, AI agents and an innovation roadmap?
“We’re clearly in front there,” he added. “We see that with our win rates, and that’s where customers like HMRC, who clearly chose to go onto our platform, going from a legacy on-prem platform that was not NiCE.”
Russell added that if a customer chooses to break things apart — CCaaS platform as distinct from their AI platform, distinct from their workforce management, “we stand up really well. But when they look at the combined offer of a unified platform and the benefits of it, we stand apart. And that’s the way that we we go to market and compete on those.”
Related Article: NiCE Makes Its Move at NiCE World 2026: Agentic AI Is Now the Architecture
What Happens When a Customer Wants an AI-Native Point Solution Instead of NiCE?
One analyst pushed on a scenario every CX leader recognizes: the AI decision isn’t always all-or-nothing. Plenty of buyers are already running high-profile AI-native tools for specific jobs — call deflection, voice agents — alongside their existing CCaaS stack. NiCE was asked how it thinks about coexisting with those tools rather than getting displaced by them.
The question sparked one of Russell’s longest answers on the call, and he leaned on scale and data, not exclusivity:
“When you take just a pure AI story alone, everyone sounds pretty similar,” Russell said. “I find it remarkable.” It’s the same conversation we had with Philipp Heltewig, chief AI officer of NiCE, back in Orlando in June at NiCE World.
Creating voice agents on Cognigy “takes seconds. It’s this is not a difficult activity.”
AI Natives Can’t Keep Up With Our Scale
So why NiCE? It’s the only platform that can run a hybrid workforce at scale, according to Russell. It’s the only platform that has a best-in-class AI capability. “We also replaced an incumbent AI-native solution at a large insurance company after winning a competitive evaluation against an AI-native solution and a large enterprise software platform.” he said, adding, “when you want to orchestrate it with the millions of voice interactions, the digital channels, the human workforce, you can do it in an interoperable way without any integration, without any latency.”
NiCE wants to interoperate within a competitive ecosystem. CCaaS integrates “beautifully” with other AI solutions, LLMs and open source platforms.
“The thing that we have the most that everybody needs that is native to our platform is our data,” Russell said. “You can’t run an AI platform without all the knowledge of all of the intents, the interactions, all the things that happens on voice is a requirement for any of those AI natives to be able to have an insightful way of being able to do containment, deflection, and others. So, our advantage is we can provide that natively with AI agents, human agents, interoperable platform, all within the data platform that is easy for our customers to use and deploy.”
Translation: NiCE isn’t asking every customer to rip out an AI-native tool they already like. It’s betting that whoever controls the underlying interaction data — the voice and digital history a contact center generates every day — becomes the platform everything else has to plug into. Whether that’s true depends on how much a given CX leader trusts a single vendor to hold that data versus keeping it portable.
Is NiCE Still in the UCaaS Business, or Not?
A Piper Sandler analyst asked about the newly expanded RingCentral partnership: NiCE will now resell RingCentral’s UCaaS product, while RingCentral continues offering NiCE’s CXone platform. Given NiCE had floated its own UCaaS ambitions in past years, why hand that business to a partner now?
“Strategic partnerships don’t mean exclusive,” Russell said. “We have many customers that would like a capability that we can provide from NiCE, but it’s also recognition that many customers want a world class UCaaS platform and coexist and buy that in coexistence with their CCaaS or their CX AI platform, and we want to be active in promoting that.”
That’s a real answer, and it’s a real signal: NiCE wants to own the AI and contact center layer, and it’s fine routing UCaaS demand to a partner instead of building it out itself. Worth noting for any CX leader who assumed NiCE was building toward a single-vendor UCaaS-plus-CCaaS bundle — that ambition, if it existed, has been quietly set aside in favor of “coexist.”
Related Article: NiCE Launches Dedicated AI Innovation Lab to Push Agentic CX to Enterprise Scale
Big Picture: The AI-in-CX Story NiCE Is Selling
Analyst-poking aside, every AI-native CX vendor is telling some version of the same story right now: the contact center is no longer a place where AI assists a human. It’s becoming a place where AI resolves the interaction, and a human only gets pulled in when something goes wrong.
NiCE CEO Scott Russell put it plainly. NiCE is no longer only orchestrating interactions but automating intent to resolution with agentic AI embedded across the entire service journey. At NiCE World in June, the company went further, describing agentic AI as “native at the core” of its platform rather than a feature bolted on top. Competitors are saying nearly the same thing this year, including 8×8, which reported its own AI-adoption surge the day before NiCE’s print.
“Enterprises achieving measurable business outcomes with AI in production,” Russell said on today’s call.
How the NiCE and AI Pitch Has Changed Since 2024
NiCE’s language about what AI is supposed to do inside a contact center has shifted meaningfully across three straight years of conference keynotes and earnings calls — from Barak Eilam’s “holy grail” framing in 2024 to Russell’s “architecture” framing today.
Dimension202420252026Who’s telling the storyCEO Barak EilamCEO Scott RussellCEO Scott RussellWhat they call the platformCXone MpowerCXone MpowerCXone (now fully Cognigy-native)The one-line pitch“The holy grail of CX” (said by former exec Barry Cooper) — human and AI working side by side”No longer only orchestrating interactions, we’re automating intent to resolution”Agentic AI is now the architecture — one intelligent operating modelWhere humans fit inAgents “orchestrate” AI, monitoring dozens of AI and human agents at onceAI and humans work side-by-side, managing full workflows togetherAI resolves independently; humans get pulled in only when it matters mostThe proof point they led with”Reverse prompting” — AI asks agents for a teachable moment and learns from the exchangeA $100 million-plus deal with an unnamed European government agencyHMRC, via Capgemini — called their biggest CXone and Cognigy deal ever
Read straight down the table, the argument has moved in one direction: 2024 was AI helps the human. 2025 was AI works alongside the human and deals prove it. 2026 is AI works instead of the human, and the architecture proves it. The job description for a contact center agent keeps shrinking in each retelling.
How 8×8 Tells a Similar Story, Differently
NiCE isn’t the only CX-adjacent vendor reporting an AI surge this earnings season. 8×8 released Q1 FY27 momentum metrics on Aug. 4, the day before NiCE’s print. On their face, the numbers are more dramatic: AI solution adoption more than doubled year over year (up 121%), 8×8 Engage’s customer base grew 247%, and usage-based revenue — the bucket that includes AI, communication APIs and digital channels — grew 63% year over year.
8×8’s Chief Product Officer Hunter Middleton framed it this way: “AI adoption more than doubled, and it’s showing up in live customer conversations, not pilots. We’re investing on both sides: the AI our customers deploy, and the infrastructure underneath that routes, processes and governs every interaction in real time. That combination is what lets them run AI-first CX at enterprise scale, and the numbers reflect it.”
Have a tip to share with our editorial team? Drop us a line: