NiCE Ltd: An AI Value Play Trading at 11x Earnings
Long-time lurker doing his first post. Formulated by AI, but thoughts are my own :)
TL;DR: NiCE provides customer service software to enterprise customers. Customer service will be heavily impacted by AI. Budget will shift from humans to software. The market is pricing NiCE as if AI will destroy its business. My thesis is that AI will destroy the old seat-based economics of customer support software, but NiCE may be one of the few incumbents capable of capturing the new economics.
A few numbers first
- Revenue grew from roughly $900m to almost $3bn over the last decade (~12% CAGR)
- The business operates at 20%+ net margins
- AI ARR is growing ~66% YoY
- The stock trades at roughly 11x earnings
The traditional business
Every company has a customer service department to deal with problems customers have with its products or services.
Historically, companies solved this problem with large amounts of headcount. Most support agents are “first-line support”. They work through relatively structured processes and handle recurring issues. Cases that don’t fit the playbook are escalated to specialists.
These employees work inside a software platform that:
- Aggregates customer and company data
- Routes cases to the right people
- Allows support agents to take actions
- Tracks performance and workflows
NiCE is one of the companies providing such software.
The AI disruption
Chatbots have existed for years.The difference is that old chatbots mostly relied on simple if/then logic and generally provided a terrible customer experience. LLMs are fundamentally different. Customer support is largely a language problem, and LLMs are very good at language.
There are two ways AI is being deployed today:
a) AI acts as a co-pilot, reading customer messages and drafting responses for human agents. b) AI responds directly to customers and increasingly takes actions by itself.
Today, fully autonomous AI is starting to be used for repetitive cases. More complex cases are still routed to humans.
I work at a software scale-up where we outsource customer support. Customer support makes up roughly 20% of our total headcount. Since introducing our AI chatbot, it now handles around 80% of incoming support requests.
We expect support headcount to decline materially over time as a result. The shocking part is that for repetitive cases, AI responses are often not just cheaper and faster, but actually better. An AI agent once responded to a customer that had just lost his mother. The response was empathetic but insisted that the invoice still had to be paid.
What this means for the industry
I am convinced that a large proportion of customer support organizations will reduce headcount by 50% or more over time, particularly for process-driven Tier 1 support.
I don’t think this happens overnight. But within 5 years, I expect AI-powered support to be the default experience for many customer interactions.
The biggest change resulting from this will be the shift from seat-based pricing towards usage-based and outcome-based pricing.
Established players such as Genesys, Five9, Talkdesk and NiCE all seem to have recognised this existential threat and are aggressively investing in AI. At the same time, there are AI-native start-ups growing incredibly quickly.
The question, as always, is whether incumbents can adopt the new technology faster than start-ups can crack distribution. My hypothesis is that start-ups will do well in the mid-market, but will struggle more with large enterprises.
Why NiCE
Along with Genesys, NiCE has some of the deepest penetration into large enterprises.
Their strongest vertical is financial services, but they are also present in healthcare, telecoms and government.
These deployments are deeply integrated into company processes and often take years to implement. That creates meaningful switching costs.
As a result, the easiest way for enterprises to adopt AI is simply to extend the platform they already use. Of all incumbent providers, NiCE appears to be one of the most aggressive investors in AI.
Their numbers support this:
- More than 10% of revenue is AI-related
- AI ARR is growing roughly 66% YoY
- Management claims AI is included in essentially all new enterprise deals
- NiCE also acquired Cognigy, one of the more interesting AI-native companies in the space. This gives them technology they can cross-sell into an already existing customer base.
I also like their go-to-market strategy. Existing customers can shift spending from seats toward AI usage without immediately increasing the total contract value, making adoption easier inside large organizations.
The transition phase
The market understands that contact-center software may be more exposed to AI disruption than many other SaaS categories.
The key question is: Will AI revenue grow faster than seat revenue declines?
Recent earnings suggest that the transition is already creating pressure. Revenue continues to grow, but profitability has come down.
My interpretation is that this reflects AI investment as well as customers gradually shifting spend toward lower margin AI products given the mentioned contract budget shifts.
The transition will almost certainly be messy, especially on pricing. Price too aggressively and customers constrain usage. Price too conservatively and margins collapse.
The bull case
Historically, customer support budgets were mostly labor budgets. If AI performs an increasing share of customer support work, some of that labor budget should migrate into software spend.
The key question is whether NiCE can capture enough of that budget shift. If AI ends up doing significantly more work than human agents, software vendors should eventually capture a larger share of the economic value created.
I believe expanding switching costs will further increase stickiness and pricing power, as AI systems add an additional layer of customisation that would have to be ripped out.
If NiCE successfully navigates the transition, it could continue its history as a compounder and its P/E ratio would be re-rated as a consequence (historically it was >40, vs. 11 PE ratio now).
I also think the downside is more limited than it may initially appear. Even if NiCE struggles to monetize AI as effectively as I expect, customer support is not disappearing. A substantial share of interactions will likely remain human-operated for a long time, particularly in regulated industries.
Those companies will still need routing, reporting, workforce management, compliance and all the other infrastructure surrounding customer support. In other words, you are buying an existing profitable (but contracting) software business and getting an AI bet at a good price on top.
Risks
The biggest risk is that AI creates more value for customers but less value for software vendors. If seat revenue disappears faster than AI revenue grows, the entire industry’s revenue pool could shrink.
It is outside my circle of competence to judge whether NiCE has the best technology, but some users dislike the software and complain it is difficult to configure. If anyone here has hands-on experience with NiCE, Genesys, Five9, Talkdesk or the newer AI-native players, I’d love to hear your perspective.
Also worth mentioning: the new CEO, NiCE has expanded partnerships with companies such as AWS and Salesforce. This helps distribution, but may also reduce lock-in and put pressure on margins.
Finally, the company is headquartered in Israel, which introduces geopolitical risk.
Conclusion
The market is pricing NiCE as if AI will destroy its business. That may happen. If seat-based revenue disappears faster than AI revenue grows, the stock probably deserves to be cheap.
My view is different. I think AI will destroy the old economics of customer support, but I also think a meaningful share of customer support budgets will migrate from labor to software. If that happens, the winners are unlikely to be random start-ups. They are more likely to be the companies that already sit at the center of large enterprise customer support operations.
NiCE is not the only company pursuing this opportunity, and there are plenty of risks. But a company with a decade-long track record of ~12% revenue growth, 20%+ net margins, no meaningful debt, AI ARR growing ~66% and one of the strongest enterprise distributions in the industry trading at roughly 11x earnings strikes me as an interesting setup.
Note: NiCE also has a financial monitoring business that I excluded from this analysis for simplicity’s sake.