| Constellation Research

The Shifting Sands of AI: Why Enterprise Leaders Need to Look Beyond OpenAI

Media Name: sand-dunes-cc0-pexels-francesco-ungaro-998657-0.jpg

Distinguished Chair of the Accelerator & Principal/CEO
Stimson Center & LeadDoAdapt Ventures

Business Insider named Dr. David A. Bray one of the top “24 Americans Who Are Changing the World” and he was named a Young Global Leader by the World Economic Forum in 2016. He has served in a variety of leadership roles in turbulent environments, including bioterrorism preparedness and response from 2000-2005. In 2019, he was invited to give the AI World Society Distinguished Lecture to the United Nations on UN Charter Day. In 2025, he was invited to be an expert witness before a public Congressional hearing on AI to help inform AI policies. In 2026, the Marconi Society selected him as inaugural winner for global leadership and Excellence in AI. ...

R "Ray" Wang

Principal Analyst and Founder
Constellation Research

R “Ray” Wang is the CEO of Silicon Valley-based Constellation Research Inc. He co-hosts DisrupTV, a weekly enterprise tech and leadership webcast that averages 50,000 views per episode and blogs at www.raywang.org. His ground-breaking best-selling book on digital transformation, Disrupting Digital Business, was published by Harvard Business Review Press in 2015. Ray's new book about Digital Giants and the future of business, titled, Everybody Wants to Rule The World was released in July 2021. Wang is well-quoted and frequently interviewed by media outlets such as the Wall Street Journal, Fox Business, CNBC, Yahoo Finance, Cheddar, and Bloomberg. Short Bio R “Ray” Wang (pronounced WAHNG) is the Founder, Chairman, and Principal Analyst of Silicon Valley-based Constellation Research Inc. He…...

A Rapidly Evolving Landscape; OpenAI's Disappearing Moat

We've been watching the generative AI landscape transform at breathtaking speed, and what concerns us most is how quickly the narrative around OpenAI has shifted from "unassailable market leader" to "company facing existential challenges." As leaders who have spent our careers at the intersection of technology, policy, and enterprise strategy, we believe that organizations making multi-million dollar AI investments need to understand the broader context beyond the marketing hype.

The concept of a "moat" in business refers to sustainable competitive advantages that protect a company from competitors. OpenAI's initial moat was built on first-mover advantage, technical superiority, and massive funding. All three pillars are now showing significant cracks.

Microsoft—OpenAI's primary backer—has began testing outside models from xAI, Meta, and even Chinese company DeepSeek. Simultaneously, Apple appears to be reconsidering its OpenAI partnership, now engaging with Google about Siri integration. These moves by two of the world's most valuable companies signal serious concerns about OpenAI's trajectory.

The technical superiority argument is also collapsing. OpenAI's rushed GPT-4.5 release shows a 30% error rate—significantly worse than both Anthropic's Claude 3.7 and xAI's Grok3. When your core product is underperforming relative to competitors, enterprise customers take notice.

Competition Is Intensifying; The Economics Don't Add Up

While OpenAI struggles, competitors are gaining momentum. Anthropic secured a $3 billion investment from Google and released Claude 3.7, which many consider technically superior to OpenAI's offerings. Elon Musk's xAI launched Grok3 with impressive deep research capabilities. Even OpenAI's former CTO, Mira Murati, launched Thinking Machines Lab and raised $2 billion at a $9 billion valuation in just two weeks.

And we can't ignore developments from China. Within the last few weeks,they announced what they described as the world's first fully autonomous AI agent, called Manus. Unlike some overhyped Western announcements, Chinese AI capabilities have generally delivered on their promises. This represents both competitive and geopolitical considerations for enterprise leaders.

The financial picture is equally concerning. OpenAI is reportedly burning through $1 billion monthly and could lose up to $44 billion by next year. Sam Altman himself admitted they lose money on every $200/month ChatGPT subscription. Their recent announcement of enterprise offerings priced between $2,000-$20,000 monthly appears to be a desperate attempt to stem these losses.

This pricing strategy reveals a company pivoting toward enterprise customers out of necessity rather than strength. But this market is already dominated by Microsoft, Amazon, and Google, who have decades-long relationships with Fortune 500 companies. OpenAI faces an uphill battle against entrenched competitors with deeper pockets and broader offerings.

Despite the recent headline-grabbing $40 billion funding round that catapulted OpenAI's valuation to $300 billion and reports that the company's revenue has grown by 30% in three months, the company still doesn't expect to break even until 2029—four years from now! This timeline raises serious questions about the sustainability of their business model, especially as they continue to burn through cash at an alarming rate.

In a telling strategic pivot, OpenAI has also announced plans to launch an open-weights reasoning model that developers can run on their own hardware. This represents a significant departure from their closed system subscription model and suggests an acknowledgment that their current approach may not remain competitive in the long term. This move appears to be a course correction in response to mounting pressure from both open-source alternatives and competitors offering more flexible deployment options.

Strategic Implications for Enterprise Leaders

For CEOs, CTOs, CIOs, and CMOs, these developments necessitate a more sophisticated approach to AI strategy. The days of simply "partnering with OpenAI" as a complete AI strategy are over. We believe enterprise leaders need to consider:

Geopolitical factors: How will US-China tensions affect your AI supply chain? What regulatory frameworks are emerging in different regions?
Economic sustainability: Are your AI partners financially viable for the long term? What happens if they significantly raise prices or pivot their business models?
Technical diversification: How can you build an AI architecture that isn't dependent on a single provider?

Enterprise clients can implement what we call a "multi-modal, multi-model" approach. This means leveraging different AI models for different use cases and maintaining the flexibility to switch providers as the landscape evolves. The companies that will win in the AI era aren't those that pick the "right" vendor today, but those that build adaptable AI architectures.

OpenAI's current valuation approaching $300 billion seems increasingly disconnected from economic reality. While they deserve credit for catalyzing the current AI revolution, enterprise leaders need to recognize that we're entering a new phase where multiple players will drive innovation.

The next 18 months will be critical. We'll see consolidation among smaller AI companies, continued heavy investment from tech giants, and potentially surprising moves from nation-states viewing AI as critical infrastructure. Enterprise leaders need to stay informed not just about the technology, but about these broader market and geopolitical dynamics.

The bottom line for enterprise leaders: your AI strategy needs to be as sophisticated as the technology itself.

Look beyond the hype, consider the full spectrum of factors at play, and build flexibility into your approach. We believe the latest "wave" of the current AI revolution is just beginning, and the winners will be those who navigate its complexities with clear-eyed strategic thinking.

Data to Decisions Future of Work Innovation & Product-led Growth Marketing Transformation New C-Suite Tech Optimization Chief Analytics Officer Chief Data Officer Chief Digital Officer Chief Executive Officer Chief Information Officer Chief Privacy Officer Chief Procurement Officer Chief Product Officer Chief Supply Chain Officer Chief Technology Officer

Read more about The Shifting Sands of AI: Why Enterprise Leaders Need to Look Beyond OpenAI

Qualtrics CEO Jason Maynard on the Future of Experience Management

R "Ray" Wang

Principal Analyst and Founder
Constellation Research

For years, experience management has been built around a relatively simple model: listen, understand, act. Companies collect feedback. They measure satisfaction. They map customer journeys. They identify problems and then try to fix them. But AI is changing what comes next.

I recently sat down with Jason Maynard, CEO of Qualtrics, to talk about the future of experience management and how AI is changing the way organizations understand, predict, and deliver experiences. The biggest shift may be moving from looking backward to predicting the future.

From the rearview mirror to simulation

Maynard describes the evolution of experience management as a move from understanding what happened to predicting what could happen next.

His analogy is Formula 1 racing. Teams don't simply look at what happened in the last race. They use massive amounts of data and simulation to test strategies and anticipate different outcomes before race day.

Maynard believes organizations will increasingly do the same with their customer and employee experiences. Companies could simulate changes to pricing, product packaging, customer engagement, employee interactions, and other experience outcomes before making those changes in the real world.

That could make testing both cheaper and more precise. Instead of waiting for an experience to happen and then measuring the result, companies can increasingly ask:

What happens if we change this? And then test it before making the decision.

The journey map is becoming an experience loop

One of the more interesting ideas from the conversation is the shift from journey maps to experience loops.

Journey maps are static. They describe a path through an experience and often require people to manually determine what happens next. Experience loops are different. They are continuous. They can adapt based on new information, simulate different paths, and keep learning from what happens.

That matters because customers don't experience companies in neat, linear journeys. They move between channels, products, people, and systems. Their needs change. Their behavior changes. Their context changes. AI makes it possible to adapt continuously to those changes rather than relying on a journey map that eventually ends up sitting in a drawer.

Experience may be the differentiator AI can't easily copy

We also talked about why experience is becoming even more important. Products can be copied. Pricing can be copied. Technology can be replicated.

Experience is harder. The challenge is making that experience personal at scale. A customer doesn't want an organization to simply know their data. They want it to remember their preferences, understand their context, and respond accordingly.

That requires more than another application. It requires organizations to connect the data and signals that already exist across the enterprise.

AI doesn't eliminate the data problem

There is an obvious catch. Simulation is only as good as the data behind it.

Maynard argues that Qualtrics' advantage comes from more than two decades of experience data and its ability to organize that information through an experience ontology. The bigger enterprise challenge is connecting that experience intelligence with the rest of the organization's data.

Most companies have already accumulated massive application and data sprawl. SaaS multiplied the number of point products, and each one created another source of information. AI can reason across those systems, but organizations still need to know which data matters, what it means, and where the gaps are.

The problem isn't simply having more data. It's having the right context.

Decision velocity becomes the new advantage

Perhaps the biggest implication is organizational. AI isn't just changing the technology stack. It's changing how quickly companies can make decisions. If organizations can simulate more options, learn from more signals, and continuously close the loop between action and outcome, the distance between strategy and execution starts to shrink.

That is what Maynard calls decision velocity. And it could become one of the most important measures of AI maturity. The winners won't necessarily be the companies with the most AI. They'll be the companies that can use AI to learn faster, decide faster, and adapt faster.

The next era of experience management isn't just about listening. It's about simulating, predicting, and continuously delivering better experiences continuously.

Next-Generation Customer Experience Tech Optimization Data to Decisions Customer Experience Data to Decisions AI Agentic AI Chief Executive Officer Chief Digital Officer Chief Technology Officer

Read more about Qualtrics CEO Jason Maynard on the Future of Experience Management

Is Meta an Enterprise Player? + Enterprise Rebranding & AI Architecture | CRTV Episode 140

Hannah Mason

Marketing Director
Constellation Research

Hannah Mason is a marketing director at Constellation Research. She oversees brand positioning, creative direction, and content development across digital platforms, events, and campaigns. Her work spans branding and design—including graphics, sales collateral, and event presentations—along with website content management, newsletters, and Constellation Insights. Mason also manages surveys and produces Constellation’s video collateral, including ConstellationTV. Her role ensures that the Constellation brand is communicated with clarity, creativity, and consistency across all channels. Contact Hannah at [email protected]. ...

Esteban Kolsky

Chief Distiller, Board Advisor
Constellation Research

Esteban Kolsky spent the first part of his professional career working on the customer service front line and in management roles, culminating in early implementations of first-generation CRM tools. In 2000, he became a research analyst at Gartner, focused on CRM, customer service, customer experience management, enterprise feedback management, and voice of the customer, among topics that later became the basis for the current field of CX. During his tenure at Gartner, Kolsky assisted vendors and Global 2000 organizations with their customer strategies. He left Gartner to focus his career on strategic advisory services and influenced all the major software vendors, as well as many smaller vendors, in the CX industry and advised executive leadership teams across industries and countries…...

Enterprise AI is entering a new phase. The question is no longer simply what AI can do. Enterprises are now grappling with whether they can trust it, how to train it, what infrastructure they need, and how quickly they can adapt as the evidence evolves.

Episode 140 of ConstellationTV brings those questions together through four conversations.

Can Meta make the jump to enterprise AI?

Meta is making a bigger push into enterprise AI, but the move raises an obvious question: Does Meta have what it takes to win the enterprise?

Constellation analysts Liz Miller, Larry Dignan, and Esteban Kolsky debate Meta's enterprise ambitions, from trust and governance to talent and go-to-market experience. The company has plenty of AI capabilities and data, but enterprise buyers demand a different level of confidence.

The debate also looks at Meta's Muse AI shopping ambitions and the broader shift toward AI-powered shopping assistants, where consumer behavior could create a very different opportunity for the company.

AI needs more than data

Liz Miller then sits down with Johann Wrede to discuss Auros, the newly rebranded company formerly known as UserTesting. The bigger story isn't the name change. It's the changing role of human intelligence in AI.

Johann argues that AI reflects the data and knowledge humans give it. As the easy problems become increasingly solved, the next challenge is capturing expertise and experience that can help models and agents perform at a much higher level.

That means moving beyond simply asking whether an AI system produces the right outcome. Enterprises also need to understand how the experience feels, whether the system behaves appropriately, and whether human intelligence is being incorporated throughout the AI lifecycle.

Cheaper AI could mean more infrastructure

Next R "Ray" Wang talks with CoreWeave CEO Mike Intrator about the infrastructure powering the AI economy.

CoreWeave's thesis centers on specialized AI clouds rather than generalized infrastructure. The company sees AI workloads as requiring a more purpose-built approach, with the “minivan versus F1” analogy illustrating the difference.

The conversation then turns to token economics. As the cost of AI falls, demand can rise rather than decline. That means broader AI adoption can continue driving demand for compute, including older infrastructure that remains economically useful for different workloads.

The conclusion: more AI adoption means more demand for chips, power, land, and capital.

The enterprise needs to move faster

Esteban Kolsky closes the episode with his October Enterprise Technology Intelligence update.

The key shift is from asking whether enterprises can deploy technology to asking how quickly they can change what they have deployed when the evidence changes.

Budgets are becoming more dynamic. Decisions that once took months can happen in weeks. AI is creating productivity gains, but the harder question is what enterprises do with the capacity those gains create.

At the same time, synthetic data, more models, more clouds, and more providers are creating new options. But options only create resilience if enterprises can actually use them across their existing architecture and processes. As AI agents gain more authority, enterprises will also need new approaches to control and access during execution.

That's the thread connecting Episode 140: AI capability is advancing quickly. The competitive advantage will come from how quickly enterprises can adapt around it.

Data to Decisions Tech Optimization Future of Work Innovation & Product-led Growth Board Strategy Agentic AI AI Generative AI Machine Learning LLMs Chief Executive Officer Chief Marketing Officer Chief Technology Officer Chief Data Officer

Read more about Is Meta an Enterprise Player? + Enterprise Rebranding & AI Architecture | CRTV Episode 140

On ConstellationTV

CoreWeave'e AI Cloud: From Infrastructure to the AI Loop

R "Ray" Wang

Principal Analyst and Founder
Constellation Research

Holger Mueller

Vice President and Principal Analyst
Constellation Research

Holger Mueller is VP and Principal Analyst for Constellation Research for the fundamental enablers of the cloud, IaaS, PaaS and next generation Applications, with forays up the tech stack into BigData and Analytics, HR Tech, and sometimes SaaS. Holger provides strategy and counsel to key clients, including Chief Information Officers, Chief Technology Officers, Chief Product Officers, Chief HR Officers, investment analysts, venture capitalists, sell-side firms, and technology buyers.  Coverage Areas: Future of Work Tech Optimization & Innovation  Background: Before joining Constellation Research, Mueller was VP of Products for NorthgateArinso, a KKR company. There, he led the transformation of products to the cloud and laid the foundation for new Business Process as a…...

I'm at CoreWeave's Fully Connected 2026 in San Francisco this week, the company's first-ever global user conference, and I sat down with Susanne Seitinger, VP of Product Marketing at CoreWeave, and my colleague Holger Mueller to talk about the company's big announcement: Forge.

Forge is CoreWeave's opinionated statement on what Susanne calls the "AI loop," the full set of steps it takes to make AI genuinely useful inside an organization. It's built on what CoreWeave calls metal-to-model thinking: everything connects back to the hardware layer so you can optimize across the entire stack, not just the parts of it you can see from an application dashboard.

What struck me most is that this isn't a one-time delivery. As Susanne put it, the model delivery lifecycle is "never a one-time shot, it's continuous, a virtuous, infinite loop." You get value earlier in that loop because you aren't stuck waiting for a signal on whether something is actually working. And the variable most teams assume is driving change, the data, is only part of the story. The infrastructure underneath it, the GPUs, changes constantly too, with real cost implications for what you can afford to run and how.

The only way to know if any of it is working, Susanne told us, is to talk to your customers directly and bring that signal back into the loop. That's exactly what we saw happening on stage with customers like Caterpillar and Capital One. Caterpillar is thinking about its business over the next hundred years, and that changes what kind of cloud partner it needs. This isn't about shaving costs. It's about scaling up and building a long-term partnership around much bigger objectives.

That takes me to the five themes I keep coming back to in nearly every enterprise AI strategy conversation right now:

Margin compression. Most of the Fortune 500 are only growing in single digits. There's no room left to find that growth through cost-cutting alone.
Exponential scale. Nothing's interesting anymore unless you're doing it 10x, 100x, or 1000x better. Incremental improvement doesn't move the needle.
A mindset shift. This isn't about cost savings anymore. It's about expansion as a core strategy.
Infinite possibilities through cycle expansion. This is the one people underestimate. If you can get 2x, 5x, or 10x more "at-bats" than your competitor, more reps, more iterations, more chances to learn, that's not a cost story. That's a transformation story.
Strategy and execution becoming one thing. With infinite possibilities on the table, you don't have time to mess up. It's a game of velocity, not speed.

Holger's favorite part of the announcement was the backward compatibility CoreWeave has built in: the ability to run everything from the newest Rubin architecture back to Volta from 2017, six full GPU generations, working in concert through a standard operating model CoreWeave calls Mission Control. That consistency is what actually lets you scale. As Holger pointed out, some things have to stay consistent, or scaling simply isn't possible.

That consistency is also what makes a concept like infinite computing practical rather than theoretical. Nothing is truly infinite, Holger quickly noted, but for the practical purpose of running your business, you don't need to know exactly how much capacity exists. You use what helps your business, what you can afford, what accelerates you, and it becomes practically infinite. That reframes how you design and automate processes from the ground up: you stop asking for ten GPUs or fifty, and start asking for as many as it takes to transform the business.

The through-line across the entire conversation was partnership and composability. CoreWeave is deliberately building openness into Forge, including a partner network and search capabilities through Query Forge, because agents need to talk to the rest of the world to be genuinely useful. And the roadmap itself comes directly from customers. As Susanne said, "we built Arena because we heard that's what they need." That kind of direct signal from the field is exactly what the next phase of enterprise AI strategy has to be built on.

My takeaway: the future of cloud isn't a single switch from "build" to "buy." It's a spectrum from turnkey to build-to-suit, and for the first time, one platform shows the whole spectrum. That's the opportunity in front of enterprises right now, and it's why conversations like this one matter more than ever.

Tech Optimization Innovation & Product-led Growth Future of Work AI Agentic AI Generative AI Chief Marketing Officer Chief Technology Officer

Read more about CoreWeave'e AI Cloud: From Infrastructure to the AI Loop

A Conversation on AI for Manufacturers | R "Ray" Wang & Sanjay Brahmawar

R "Ray" Wang

Principal Analyst and Founder
Constellation Research

QAD | Redzone CEO Sanjay Brahmawar sits down with industry analyst and influencer R "Ray" Wang, founder and CEO of Constellation Research, to unpack the biggest shift in manufacturing over the last 12 months: the move from AI hype to AI outcomes. Sanjay shares what he's hearing from customers on the ground: faster product development cycles, compressed time-to-value, and AI being used to solve real manufacturing constraints. They discuss why boards are raising the bar on AI expectations, how agentic AI is stepping in to augment a shrinking manufacturing workforce, and why "exponential efficiency" is now the goal rather than incremental gains.

🔗 Learn more: https://www.qad.com

Innovation & Product-led Growth Tech Optimization Future of Work Chief Technology Officer Chief Executive Officer

Read more about A Conversation on AI for Manufacturers | R "Ray" Wang & Sanjay Brahmawar

AI Gets Cheaper. Demand Gets Bigger: Inside CoreWeave’s AI Infrastructure Strategy

R "Ray" Wang

Principal Analyst and Founder
Constellation Research

AI's economics are changing fast. Token costs are coming down. Models are becoming more accessible. Enterprise use cases are expanding. And instead of reducing the need for infrastructure, those trends could create substantially more demand for it.

That was one of the central themes in a conversation between Constellation Research founder R "Ray" Wang and Mike Intrator, CEO of CoreWeave, at Fully Connected 2026. The discussion unpacked what happens when AI moves from a frontier technology into a broader economic infrastructure layer.

AI infrastructure is not just another cloud

CoreWeave's thesis starts with a simple idea: AI workloads are different.

The company was built around specialized infrastructure designed specifically for AI, rather than adapting a generalized cloud to the demands of AI workloads.

Intrator compares the difference to that between a minivan and an F1 car. A generalized cloud can do many things well. AI infrastructure needs to be purpose-built for workloads operating at an entirely different scale.

That specialization extends beyond compute. Intrator argues that the software stack, cloud, security and other components all need to come together as one functioning system. AI infrastructure doesn't work if one piece of the puzzle is missing.

Cheaper AI could mean more infrastructure demand

One of the most interesting parts of the conversation was the economics of inference. As token costs decline, the obvious assumption is that infrastructure economics should get tougher. But Intrator sees another dynamic at work: lower costs can expand demand.

That's the Jevons Paradox at work. As AI becomes cheaper to use, more organizations can afford to use it, and existing users can expand their use.

Intrator points to the broader diffusion of AI across the economy as a key driver. Lower cost per token makes it possible to bring AI into more decisions, more workflows, and more use cases. That creates an important distinction for enterprise leaders: cheaper AI does not necessarily mean less infrastructure. It can mean more AI.

The GPU lifecycle may be longer than expected

Another important point is that not every AI workload needs the newest hardware.

As AI use cases mature, workloads can be matched to different infrastructure levels. CoreWeave has seen older GPUs remain economically valuable as customers right-size hardware for specific workloads.

Intrator points to A100 GPUs already contracted through 2029 as an example of how long the economic life of AI infrastructure can extend. That changes how enterprises should think about AI infrastructure. The race isn't necessarily about replacing everything with the newest chip as quickly as possible. It's also about matching the right infrastructure to the right workload.

The model doesn't matter as much as the infrastructure

The conversation also touched on the rise of open-weight models and lower-cost alternatives.

CoreWeave's position is relatively model-agnostic. Intrator argues that both open and closed-model ecosystems can flourish, while the physical infrastructure remains fundamental to delivering AI at scale.

That's an important distinction. The AI model layer will continue to change rapidly. The underlying infrastructure required to train, run, and serve those models remains a more persistent requirement.

AI infrastructure is moving into the enterprise

The next phase is not just about frontier labs.

CoreWeave is increasingly focused on making infrastructure usable across a broader customer base, including enterprises. Intrator cites companies such as Caterpillar and Capital One as examples of organizations that need access to AI infrastructure and the tooling to actually use it across the business.

That's where Forge comes into the conversation.The goal is to bring together the different pieces of CoreWeave's ecosystem and provide the tools customers need to build and operate AI applications across their organizations.

The implication is significant: AI infrastructure is shifting from something consumed primarily by AI specialists to something the broader enterprise needs to operate.

The infrastructure race is far from over

The conversation ended with a deceptively simple question: Where will chips, energy, land and demand be in three years?

Intrator's answer was consistent across all four: up.

His underlying argument is that demand for AI is becoming deeply rooted and sustained as the technology moves from frontier labs into enterprises and broader society. More adoption means more compute, which means more chips, power, land, and capital.

That may be the biggest takeaway from the conversation. The AI infrastructure story isn't simply about building enough capacity for today's models. It's about preparing for what happens when AI becomes cheaper, more accessible, and embedded in more of the economy.

The models will change. The workloads will change. The economics will change....But the infrastructure requirement is only getting bigger.

Data to Decisions Innovation & Product-led Growth Tech Optimization New C-Suite AI Agentic AI Generative AI LLMs Chief Executive Officer Chief Technology Officer Chief Data Officer Chief AI Officer

Read more about AI Gets Cheaper. Demand Gets Bigger: Inside CoreWeave’s AI Infrastructure Strategy

On Event Update

UserTesting Becomes Auros: Building the Human Intelligence Layer for AI

Liz Miller

Vice President & Principal Analyst
Constellation Research

About Liz Miller: Liz Miller is Vice President and Principal Analyst at Constellation, focused on the org-wide team sport known as customer experience. While covering CX as an enterprise strategy, Miller spends time zeroing in on the functional demands of Marketing and Service and the evolving role of the Chief Marketing Officer, the rise of the Chief Experience Officer, the evolution of customer engagement, and the rising requirement for a new security posture that accounts for the threat to brand trust in this age of AI. With over 30 years of marketing experience, Miller offers strategic guidance on the leadership, business transformation, and technology requirements to deliver on today’s CX strategies. She has worked with global marketing organizations to transform everything from…...

Hannah Mason

Marketing Director
Constellation Research

This just in: UserTesting is becoming Auros.

In this CR CX Convo, Constellation analyst Liz Miller sat down with Johann Wrede, CMO of UserTesting, to unpack this hot-off-the-press announcement and what it signals about where the company is headed.

This is more than a name change. Auros represents a broader vision for the role human intelligence can play in the AI era.

UserTesting and User Interviews are not going away. Johann was clear that the company remains committed to those products and the UX research community. Instead, Auros expands the portfolio and the mission beyond traditional usability testing.

From testing software to shaping AI

The shift makes sense when you consider how the interaction model is changing.

For years, UserTesting helped companies answer questions like: Does this screen make sense? Can someone accomplish what they need to accomplish?

But what happens when there is no screen?

With an AI agent, the interaction may simply be a conversation. The bigger question becomes whether the agent can deliver the right outcome and whether the experience feels right to the person using it. That requires a different kind of human insight.

Johann describes Auros as moving toward becoming part of the infrastructure that helps make AI applications better, while continuing to invest in UserTesting as an important part of that portfolio.

AI needs more than more data

One of the themes Johann and Liz kept coming back to was data.

AI reflects what it learns from. If the underlying data contains bad assumptions, incomplete perspectives or poor examples of how something should work, AI can carry those problems forward. The answer isn't simply feeding AI more information.

It's giving AI better information from people who actually have the expertise.

A radiologist brings a different level of expertise to an X-ray than someone who simply knows how to read one. A procurement expert brings a different perspective to procurement than someone who has scraped the internet for everything related to procurement.

That distinction between experience and expertise becomes increasingly important as AI moves into more specialized work.

Human intelligence moves upstream

Traditional research can sometimes feel like a final checkpoint. Build something, test it, find the problems, and iterate.

AI requires us to rethink that model. Human intelligence can become part of the process from the beginning, helping train models, validate outputs and make sure agents behave the way they are intended to behave.

That is a significant shift. We're no longer just asking people to tell us whether an AI experience worked. We're asking them to help define what good looks like in the first place.

Common sense isn't actually common

One of the best parts of the conversation was our discussion about “common sense.”

Because whose common sense?

A procurement leader in New York may bring a very different perspective than someone working in procurement for an NGO in Africa or for a manufacturing organization. Those differences matter.

If AI is going to operate across industries, cultures, and communities, we need more than a generic definition of "good". We need diverse perspectives, verified expertise and a better understanding of what good judgment actually looks like.

Johann sees that human network as a key part of the Auros vision.

From usability to human intelligence

This is ultimately what makes the Auros announcement interesting.

UserTesting built a business around understanding how humans interact with technology. Auros is taking that idea much further.

As AI becomes more capable, the human questions become harder, not less important:

Is the AI accurate?
Is it behaving as intended?
Does it understand the context?
Can we trust the people and expertise behind its training data?
And what does the experience feel like for the human on the other side?

AI may be changing the interface, but it isn't changing the importance of understanding people. If anything, it makes that understanding more important.

UserTesting is becoming Auros. The name is changing because the opportunity is getting much bigger.

Marketing Transformation Data to Decisions Future of Work AI Agentic AI Generative AI Marketing Chief Marketing Officer Chief Executive Officer

Read more about UserTesting Becomes Auros: Building the Human Intelligence Layer for AI

On CR Conversations

Why AI Agents Need Context, Not Just Data | Semantics & Context Management ShortList

Michael Ni

Vice President and Principal Analyst
Constellation Research

Michael Ni is Vice President and Principal Analyst at Constellation Research, covering the evolving Data-to-Decisions landscape—where CDOs, CIOs, and CPOs must modernize data infrastructure, integrate AI into decision-making, and scale automation to improve business outcomes. Ni’s research examines how enterprises operationalize AI, automate decision-making, and integrate data management and analytics into core business processes. He focuses on the challenges of scaling AI-driven decision systems, aligning data strategy with business goals, and the growing role of data and decisioning “products” in enterprise ecosystems. With 25+ years as a product and GTM executive across enterprise software, AI platforms, and analytics-driven technologies, Ni brings a practitioner’s perspective to…...

Hannah Mason

Marketing Director
Constellation Research

For years, enterprise data management has focused on a familiar set of questions: What data do we have? Can we trust it? What does it mean? Those questions still matter. But as AI agents move from generating recommendations to taking action, they are no longer enough.

An agent needs to understand not only what a piece of data means, but what matters now, what constraints apply, and what it is allowed or expected to do. That shift is turning semantics and context management into an increasingly important layer of enterprise AI architecture.

From trusted data to trusted decisions

Consider a simple business scenario. A sales pipeline has shrunk by 12%. The number alone doesn't tell an executive what to do. They need to understand which deals are at risk, what qualifies as pipeline, which costs matter, who has approval authority and what has changed with customers.

Human operators naturally fill in those gaps with experience and judgment. AI cannot simply assume that context. Once an AI agent can route work, escalate a customer or approve an exception, context becomes an architectural problem.

This is the fundamental shift from traditional data management to semantic and context management. Catalogs help govern data. Semantics establish meaning. Context adds the operational reality around that meaning: what matters now, which policies apply and what action is permitted.

The move from documents to served context

Retrieval systems can find the policy. An agent needs to apply it.

For example, a policy might say that orders above $100,000 require financial approval, restricted customers require compliance review and margins below 15% require commercial approval.

An agent still needs the current order value, customer status, projected margin, applicable approvals and the permissions governing what it can actually do. That is the move from document context to served context.

The goal is no longer simply finding trusted information. It is delivering the right context to support the right decision, under the right constraints, consistently and at scale.

Three layers of context infrastructure

The emerging category can be thought of in three layers:

Trusted business meaning: Can the organization establish consistent definitions, relationships and semantics?
Consistent consumption: Can people and AI systems use that meaning consistently across applications and workflows?
Active, current context: Can context stay current, carry policy, and reach the point where a decision is actually being made?

These capabilities span semantic management, connected semantics, AI grounding, policy, explainability, lineage, ecosystem integration, feedback and learning, and domain governance. The important change is that context is becoming active infrastructure, not simply documentation.

Why federated context matters

Enterprise data and knowledge rarely live in one place. Customer information may live in CRM. Business definitions may live in a data platform. Process state may live in a workflow system. Policies and permissions may sit somewhere else entirely.

That makes a federated approach increasingly important.

Actian, for example, is approaching this problem through a federated knowledge graph that connects metadata, lineage, business definitions, quality signals, policies and relationships across distributed systems without requiring the underlying data to move.

The architectural question is not whether enterprises will have distributed data. They already do. The question is whether AI can understand that distributed environment as a coherent business context.

Context has to stay alive

Maintaining context manually does not scale. The next generation of platforms will need to automate more of the work around discovering metadata and relationships, enriching knowledge graphs, monitoring quality, classifying sensitive data and managing governance workflows.

But maintaining context is only half the challenge. Context also needs to be activated.

Machine-readable data contracts, APIs, MCP interfaces and other integration mechanisms can make governed context available to analytics, applications and increasingly AI agents at runtime.

That is where this category starts to become much more consequential.

The new test for enterprise AI

The question for AI leaders is no longer simply whether their organization has a catalog, a knowledge graph or a semantic layer.

The better questions are:

Can the organization keep business meaning current?
Can AI access the context it needs at runtime?
Can policy be applied to the specific decision being made?
Can operational state become part of that context?
Can the system learn from what happened after the decision?

Ultimately, the test is simple:

Are your metadata investments documenting the business, or are they becoming something your AI can actually use?

As AI moves from answering questions to making and executing decisions, semantics and context are becoming less like supporting data-management capabilities and more like core enterprise AI infrastructure.

Data to Decisions Future of Work Data to Decisions AI Agentic AI Chief Data Officer Chief AI Officer Chief Technology Officer

Read more about Why AI Agents Need Context, Not Just Data | Semantics & Context Management ShortList

On ShortList Spotlights

Designing for Limitations: How TCS Is Using AI to Make Accessibility a Competitive Edge

Holger Mueller

Vice President and Principal Analyst
Constellation Research

Hannah Mason

Marketing Director
Constellation Research

For decades, accessibility lived in the same bucket as most compliance work: bolted on after the product was built, driven by regulation, and measured by whether complaints stopped coming in. In a recent interview, Constellation analyst Holger Mueller sat down with Shashi Bhushan and Dr. Charudatta Jadhav of TCS to make the case that this era is ending, and that AI is the reason the shift is happening now rather than another decade from now.

From Compliance to Competitive Advantage

The starting number in the conversation is hard to ignore: one out of six people globally lives with a disability. For years, that population was treated as a niche segment with limited purchasing power, worth accommodating for legal and reputational reasons but not worth building a strategy around. That calculation is changing. Shashi pointed to an estimated $1 trillion in aggregate purchasing power among people with disabilities globally, along with anecdotal evidence that companies that strategically serve this segment see substantially higher revenue outcomes.

The panel drew a direct comparison to how organizations now treat cybersecurity spending: initially driven by fear and regulation, but increasingly justified on business grounds. Accessibility, they argued, is on a similar trajectory, just a few years behind. As Shashi put it, once the realization sets in that this is a one-out-of-six customer base rather than an act of goodwill, board conversations shift from good intentions to competitive strategy.

Designing for Limitations, Not Personas

One of the more interesting reframes in the conversation was TCS's move away from traditional persona-based accessibility design. The conventional approach starts with a defined disability persona and designs around its known requirements, useful, but inherently reactive and limited to whatever personas were considered upfront.

TCS's alternative, which Dr. Charu called "designing for limitations," starts instead from the limitation itself: sensory, environmental, or ecosystem-level constraints. Rather than asking "how do we accommodate this persona," the question becomes "how do we solve for this class of limitation." Dr. Charu noted that this approach tends to produce a wider canvas for innovation, and frequently ends up improving the experience for people outside the original target group entirely, a pattern accessibility work has produced repeatedly produced: features built for a specific limitation often end up mainstream. Screen readers, originally built for visually impaired users, are a direct ancestor of the voice assistants now sitting on millions of kitchen counters.

Where AI Changes the Equation

The panel's most forward-looking material centered on what Dr. Charu called "egocentric models," AI systems trained specifically on accessibility-related data to understand how people with different sensory or physical constraints actually perceive and navigate the world. Unlike general-purpose large language models trained broadly on the internet, these models are purpose-built to understand, for example, how a visually impaired person constructs a mental map of a physical or digital space.

Dr. Charu extended this into a compelling long-term vision: an AI-powered "digital companion" that accompanies a person across every experience, banking, retail, travel, and beyond, retaining memory of preferences, context, and prior interactions throughout that person's life. Rather than solving accessibility one interaction at a time, the goal becomes eliminating the friction of the limitation itself from the equation entirely. Shashi, drawing on his own lived experience navigating vision-related accessibility challenges, described the shift plainly: the goal is a world where technology adapts to the human, rather than the human adapting to the technology.

The Road Ahead

Looking toward 2030, the panel pointed to two converging forces: AI's growing ability to directly enhance human capabilities, and its growing ability to build genuinely equitable systems around those capabilities. Their closing framing captured the throughline of the conversation well: enterprises built around accessibility in the AI era will need to be, in Shashi's words, SMART, sustainable, meaningful, accessible, resilient, and trusted, with technology functioning as an invisible enabler rather than a visible workaround.

The larger argument running through the discussion isn't really about accessibility as a category. It's about where the next decade of competitive differentiation in enterprise technology is likely to come from, and the panel makes a clear case that the answer includes a market segment most organizations have historically underpriced.

Future of Work Tech Optimization New C-Suite Digital Transformation Agentic AI AI GenerativeAI Chief Information Officer Chief Technology Officer Chief Human Resources Officer Chief AI Officer

Read more about Designing for Limitations: How TCS Is Using AI to Make Accessibility a Competitive Edge

On CR Conversations

On-Prem's Comeback, Qualtrics' New Playbook, and Why AI Needs Context, Not Just Data

Hannah Mason

Marketing Director
Constellation Research

Enterprise AI is outgrowing the model and the chatbot. That's the thread running through Episode 139 of ConstellationTV, where the conversation keeps landing on the same question from different angles: once AI has to actually operate inside the business, across infrastructure, customer experience, and decision-making, what does it need to work reliably?

The Big Debate: Is On-Prem AI Actually Making a Comeback?

Mike Ni moderated a debate between Holger Mueller and Esteban Kolsky on one of infrastructure's oldest questions, reopened by AI: does it run on-prem or in the cloud?

Mueller pointed to strong recent hardware numbers from Dell and HP as evidence of a real on-prem renaissance, but argued the pull toward public cloud is still structural. His analogy: running AI on-prem today is like a company generating its own electricity a century ago. It's possible, but the economics of elasticity eventually win. He introduced a concept he's calling "GPU gravity," the idea that faster GPUs will pull workloads toward them the same way data gravity pulls more data together, because in agent-to-agent interactions, the agent running on faster infrastructure simply makes better decisions, faster, more often.

Kolsky pushed back with an inference-economics argument: for steady, predictable, high-volume AI workloads, owning the infrastructure can be cheaper than the public cloud, and mainframes are seeing renewed relevance because of their CPU management strength. Ni added the practical wrinkle, that CIOs can't easily justify large capital expenditure to a board when workloads are still in experimentation mode, and that latency from data routing is a real cost most people underweight.

The debate didn't resolve into a single answer, and it wasn't supposed to. All three converged on hybrid as the near-term reality, but disagreed on the tilt: Mueller sees the long-term center of gravity moving to public cloud, Kolsky sees a genuine opening for on-prem economics, and Ni framed the real strategic question as workload routing, balancing quality, latency, data access, policy, and cost, rather than picking one mode and committing to it.

Qualtrics' New Chapter: From Measuring Experience to Simulating It

Liz Miller, Mike Ni and R "Ray" Wang sat down with the Qualtrics team in Salt Lake City following the company's recent launch event, and the shift they described is meaningful. Experience management has historically meant deciding on an experience, pushing it out, and measuring the response after the fact. Qualtrics' new direction moves that earlier: simulate the experience against synthetic panels, predict the outcome, and only then act, rather than running hundreds of live experiments against real customers.

Wang framed the shift around three ideas: simulate, trust, and outcome. The simulation piece lets teams test far more granular audience segments than traditional A/B testing ever allowed, down to sub-sub-personas rather than broad 50/30/20/10 splits. But the more interesting shift is philosophical. Miller pointed out that CMOs have spent years assuming they control the customer experience, when the customer has always been the one actually managing it. Qualtrics' new positioning directly acknowledges this, aiming to deliver "tailored" experiences that adjust in real time based on context, time, location, journey stage, even physiological signals, rather than a single predetermined path.

The two also unpacked "ontology" in plainer terms: it's the structured relationship between the signals a system tracks and what those signals actually mean for the experience being delivered. That structure, built on Qualtrics' roughly 20 years of experience data, is what makes simulation trustworthy rather than just another guess.

Splunk's Reinvention: Making Agents Reliable, Compliant, and Cost-Aware

Splunk used its recent conference to lay out a broader reimagining of itself as it deepens integration with parent company Cisco. Two announcements stood out to Constellation analyst Chirag Mehta. The new Observability Studio focuses on tokenomics, directly addressing a problem enterprises are already facing with agents in production: keeping them reliable, compliant with policy, and within budget. And Splunk is extending into security operations centers, where AI agents can now handle detection, investigation, response, and governance tasks directly.

The Cisco relationship is providing Splunk with meaningful distribution, with existing Cisco customers now able to discover Splunk's capabilities through products such as Cisco Command Center and AI Canvas. The framing throughout: Splunk's core strength was always turning data into answers, and that strength now extends to monitoring the behavior, security, and economics of AI agents themselves, not just the systems around them.

New ShortList: Semantic and Context Management

Mike Ni closed the episode by introducing Constellation's newest ShortList, covering semantic and context management, and made the case for why it deserves to be its own category rather than a subset of data management.

His framing: most executives can't cleanly define "enterprise context," but they know immediately when it's missing. A dashboard showing a 12% pipeline shrink isn't actionable on its own; a human operator fills the gap with judgment and experience. AI doesn't have that judgment, which turns missing context into an architectural problem the moment agents start taking action rather than just generating insight.

Ni drew a clear line between data management's traditional questions (what data do we have, can we trust it, what does it mean) and the newer layer semantic and context management adds: what matters right now, and what is an agent allowed or expected to do about it. He evaluated the category across nine capabilities, from semantic management and AI grounding to policy-aware context and explainability, and spotlighted Actian as a vendor worth watching.

Actian's approach centers on a federated knowledge graph that connects metadata, lineage, definitions, and policy across systems without requiring organizations to physically move their underlying data, paired with a knowledge graph that captures relationships and dependencies in a form AI systems can actually traverse, not just search. Ni also highlighted Actian's move toward "activated" context: machine-readable data contracts and support for standards like MCP that let governed context reach analytics tools, applications, and AI agents directly at runtime.

His closing question for AI leaders is the one worth sitting with: are your metadata investments just documenting the business, or are they becoming something your AI can actually use?

The Throughline

Every segment in this episode circles back to the same tension: enterprise AI's hard problems have moved past the model itself and into the infrastructure, context, and operating choices that surround it. Where workloads run, how experience decisions get made, whether agents stay reliable and cost-aware, and whether AI has the context it needs to act rather than just respond — these are the questions defining the next phase of enterprise AI adoption.

Tech Optimization Data to Decisions Digital Safety, Privacy & Cybersecurity Marketing Transformation Cloud Data to Decisions AI Agentic AI cybersecurity Chief Information Officer Chief Technology Officer Chief Marketing Officer Chief AI Officer

Read more about On-Prem's Comeback, Qualtrics' New Playbook, and Why AI Needs Context, Not Just Data

On ConstellationTV

Subscribe to

Results

The Shifting Sands of AI: Why Enterprise Leaders Need to Look Beyond OpenAI

A Rapidly Evolving Landscape; OpenAI's Disappearing Moat

Competition Is Intensifying; The Economics Don't Add Up

Strategic Implications for Enterprise Leaders

Qualtrics CEO Jason Maynard on the Future of Experience Management

From the rearview mirror to simulation

The journey map is becoming an experience loop

Experience may be the differentiator AI can't easily copy

AI doesn't eliminate the data problem

Decision velocity becomes the new advantage

Is Meta an Enterprise Player? + Enterprise Rebranding & AI Architecture | CRTV Episode 140

Can Meta make the jump to enterprise AI?

AI needs more than data

Cheaper AI could mean more infrastructure

The enterprise needs to move faster

CoreWeave'e AI Cloud: From Infrastructure to the AI Loop

A Conversation on AI for Manufacturers | R "Ray" Wang & Sanjay Brahmawar

AI Gets Cheaper. Demand Gets Bigger: Inside CoreWeave’s AI Infrastructure Strategy

AI infrastructure is not just another cloud

Cheaper AI could mean more infrastructure demand

The GPU lifecycle may be longer than expected

The model doesn't matter as much as the infrastructure

AI infrastructure is moving into the enterprise

The infrastructure race is far from over

UserTesting Becomes Auros: Building the Human Intelligence Layer for AI

From testing software to shaping AI

AI needs more than more data

Human intelligence moves upstream

Common sense isn't actually common

From usability to human intelligence

Why AI Agents Need Context, Not Just Data | Semantics & Context Management ShortList

From trusted data to trusted decisions

The move from documents to served context

Three layers of context infrastructure

Why federated context matters

Context has to stay alive

The new test for enterprise AI

Designing for Limitations: How TCS Is Using AI to Make Accessibility a Competitive Edge

From Compliance to Competitive Advantage

Designing for Limitations, Not Personas

Where AI Changes the Equation

The Road Ahead

On-Prem's Comeback, Qualtrics' New Playbook, and Why AI Needs Context, Not Just Data

The Big Debate: Is On-Prem AI Actually Making a Comeback?

Qualtrics' New Chapter: From Measuring Experience to Simulating It

Splunk's Reinvention: Making Agents Reliable, Compliant, and Cost-Aware

New ShortList: Semantic and Context Management

The Throughline