Microsoft’s
Copilot problem is increasingly becoming a brand problem, a workflow problem, and, for investors, a growth problem. When a fund manager says the product “feels like Teams” and that her firm is replacing it with
Claude, that is not just a snarky sound bite; it is a shorthand critique of how enterprise users experience Microsoft’s AI ambitions in the real world. The timing is awkward for Redmond, because Microsoft is simultaneously pushing a broader
Copilot overhaul and preparing the
E7 suite launch, while Wall Street is watching cloud growth, software multiples, and AI monetization with unusually sharp eyes.
Overview
The latest criticism of Copilot came from Bryn Talkington of Requisite Capital Management, who argued that the product feels so much like Microsoft Teams that users do not want to use it. Her firm, she said, is removing Copilot and swapping in Claude instead, a notable reversal in a market where Microsoft has spent two years trying to make Copilot the default AI layer across work software. That comment matters because it is not coming from a consumer reviewer or a one-off power user, but from a professional investor with a direct view into enterprise software buying behavior.
The broader context is that Microsoft has turned Copilot into a central pillar of its AI story. The company has woven it into Windows, Microsoft 365, Dynamics, Edge, and now enterprise governance and agent tooling. That strategy creates enormous distribution leverage, but it also risks making Copilot feel less like a breakthrough product and more like another feature inside a familiar Microsoft stack. In
some environments that is an advantage, but in others it can make the tool feel bureaucratic rather than magical.
At the same time, Microsoft is under pressure to prove that its enormous AI spend is translating into durable demand. Investor attention is shifting from launch announcements to adoption, engagement, and retention. That is where the comparison with Claude becomes more interesting: Anthropic’s product may lack Microsoft’s installed base, but it benefits from a reputation for responsiveness, cleaner interaction design, and strong performance in certain knowledge-work scenarios. In a crowded market, those perceptions can matter almost as much as raw distribution.
The timing also lines up with fresh signs that Microsoft wants to reset the Copilot narrative. The company has been highlighting new agentic features, broader model support, and a more unified enterprise packaging approach. Microsoft’s upcoming
E7 bundle, announced for May 1 availability, folds Copilot into a larger “frontier” suite that includes security, identity, and agent governance products. That is a classic Microsoft move: solve product weakness by bundling it into a platform story.
For investors, though, the issue is more immediate. Microsoft’s stock has been punished this year, and bull cases are increasingly tied to Azure growth, enterprise AI pricing, and whether the company can keep its software franchise from getting repriced alongside the rest of SaaS. This is why comments about Copilot usability resonate so strongly: if the AI layer is not loved by users, the whole monetization stack becomes harder to defend.
Why the Copilot Critique Hits a Nerve
Talkington’s complaint is powerful because it is simple. Saying a product “feels like Teams” immediately evokes the idea of software that is deeply embedded but not especially delightful. Teams is useful, but it is also associated with enterprise friction, compliance overhead, and a kind of default corporate inertia. When that label gets attached to Copilot, the subtext is that Microsoft may have built something that is technically available but emotionally resistible.
That distinction matters in AI. The best AI tools are not only accurate; they are inviting. Users need to feel that the assistant makes their work easier without forcing them into a maze of menus, permissions, and business logic. If a product feels like yet another Microsoft admin experience, then the brand itself can become a barrier to adoption,
even when the underlying technology is competent.
The Product Experience Problem
The product challenge is not limited to interfaces. It is also about how quickly Copilot can answer, how well it handles ambiguity, and how naturally it fits into daily work. Claude has earned praise in many circles for conversation quality and longer-form reasoning, while ChatGPT remains the dominant public benchmark for mindshare and usage. Microsoft, meanwhile, has to prove that Copilot can do more than sit inside existing workflows.
A company can win on distribution and still lose on delight. That is why software buyers increasingly talk about
time-to-value and user acceptance rather than just seat count. If employees avoid a tool, the enterprise ROI collapses, no matter how elegant the procurement contract looks.
- User experience now affects enterprise buying more directly than it did in older software cycles.
- Copilot’s distribution advantage does not guarantee habitual use.
- A product can be heavily licensed and still be underused.
- The “feels like Teams” critique signals friction, not just preference.
- Claude’s appeal appears to be as much about experience as about feature count.
Why Brand Perception Matters
Microsoft is in an unusual position: it is one of the world’s strongest enterprise software vendors, yet it is trying to compete in a category that is still being defined by product feel and public enthusiasm. That creates a tension between reliability and excitement. Microsoft often wins the former; competitors often win the latter.
This is why the Copilot debate is not just about one application. It is about whether Microsoft can make its AI layer feel native and indispensable rather than imposed. In
practical terms, that means the company must keep improving the product while also communicating that it is not just Office with a chatbot bolted on.
The Claude Advantage in the Enterprise AI Race
Anthropic’s Claude is benefiting from a very different market position. It does not need to be everything to everyone, and that may be part of the appeal. For users who want a focused assistant that feels responsive and coherent, Claude can seem more like a
work tool and less like a corporate platform experiment.
That does not mean Claude is categorically better. It means it may be better for specific tasks and user expectations. In enterprise settings, the buyer often cares about trust, consistency, and output quality more than novelty. If a team tests Copilot and Claude side by side and prefers Claude, that is a meaningful signal, even if it is not yet a mass-market verdict.
Focus Beats Bloat
One reason Claude is resonating is that it can feel more purpose-built. Microsoft, by contrast, is trying to integrate Copilot across nearly every product line. That breadth creates strategic value, but it can also create user confusion. People may not know whether Copilot is a search layer, a writing assistant, an agent platform, or a workflow orchestrator.
When tools do too many jobs at once, they can become harder to explain and harder to champion internally. Claude’s cleaner positioning may help it win pilots, especially in teams that care about drafting, analysis, and synthesis. In that sense, the market may be rewarding
clarity as much as capability.
- Claude benefits from a more focused product identity.
- Microsoft has to balance breadth with simplicity.
- Enterprise teams often prefer tools that are easy to explain internally.
- AI assistants are judged by output quality and interaction style.
- A narrower tool can sometimes feel more premium than a broader suite.
The Competitive Signal for Microsoft
If a fund manager is publicly saying her firm is replacing Copilot with Claude, that creates reputational risk well beyond one account. It suggests that Microsoft’s AI penetration may be deeper on paper than in practice. Competitors can use that gap to pitch themselves as the better day-to-day choice for knowledge workers.
For Microsoft, the answer is not necessarily to narrow its ambition. It is more likely to double down on integration, model flexibility, and workflow-native features. But that only works if the experience improves enough that users stop comparing Copilot to the software they already tolerate.
What the Usage Numbers Actually Tell Us
The usage figures cited around the CNBC discussion underscore the central problem. Copilot’s daily active user count is materially lower than ChatGPT’s and even below Claude’s in the cited Sensor Tower data. That does not make Copilot irrelevant, because enterprise tools often monetize differently from consumer apps, but it does suggest that Microsoft has not yet achieved the kind of broad habit formation it wants.
This is where the distinction between
distribution and
engagement becomes critical. Microsoft can place Copilot in front of millions of users through Windows and Microsoft 365, but that does not guarantee frequent use. If users open the feature occasionally and then abandon it, the engagement curve never becomes self-sustaining.
Why DAU Matters, But Not Alone
Daily active users are an imperfect metric, yet they remain useful because they measure routine behavior rather than one-time trial. In AI, habitual use is the real prize. It is what turns a prompt box into a workflow dependency.
Still, investors should not overread the headline counts. Enterprise software often monetizes through seats, bundles, and upsells, so a product can have lower DAU and still be strategically important. The key question is whether those users become sticky enough to justify the infrastructure and licensing model. That is the
real test.
Comparing Microsoft, Claude, and ChatGPT
ChatGPT’s scale reflects a broader consumer and professional audience, while Claude’s momentum suggests strong traction with power users. Copilot sits in a different lane: it is embedded in Microsoft’s work stack and therefore measured partly by attach rate, not just standalone usage. But that does not fully insulate it from comparison.
If users who can choose among tools are opting for Claude, then Microsoft has a product-quality issue to solve. If the company cannot reverse that trend, it may end up selling Copilot more like a feature bundle than a must-have assistant. That is not fatal, but it is less valuable than the vision Microsoft has been pitching.
- Copilot’s DAU profile suggests adoption friction.
- ChatGPT remains the category giant by usage.
- Claude appears to be winning some professional workflows.
- Microsoft’s embedded model helps distribution, but not necessarily love.
- Habit formation will matter more than launch-day excitement.
The E7 Suite and Microsoft’s Enterprise Reset
Microsoft’s answer to these concerns appears to be a broader enterprise reset. The company is pushing the
E7 suite, a premium package that bundles Microsoft 365 E5, Copilot, security tools, and Agent 365 into a more integrated offering. That is a strategic move designed to make AI easier to buy, govern, and deploy inside large organizations.
The logic is straightforward. If Copilot alone does not feel indispensable, then package it with the rest of the control plane and make the whole platform feel like a modern operating layer for work. Microsoft is betting that CIOs and IT leaders will prefer an integrated stack over buying scattered AI tools from multiple vendors.
Bundle Strategy as Defense
Bundling is one of Microsoft’s oldest competitive weapons. It can suppress rival point solutions, simplify procurement, and increase average revenue per user. It can also hide product shortcomings by making the customer evaluate the system rather than the app.
That said, bundle strategy only works when the underlying components are credible. If Copilot remains underwhelming, the E7 package may be perceived as a way to monetize inertia rather than a true upgrade. Enterprise customers are
much more alert to that now than they were in earlier cloud cycles.
Why Agent 365 Matters
The launch of Agent 365 is important because it shows Microsoft is not just selling a chatbot; it is trying to become the governance layer for AI agents. That is a much bigger market. If enterprises adopt multiple agents from different vendors, they will need policy, identity, monitoring, and security controls.
Microsoft is trying to own that layer before rivals do. In theory, that could turn Copilot from a product into a platform entry point. In practice, it requires trust, clarity, and adoption. If users do not like the assistant, it becomes harder to convince them to build around the platform.
Azure Is Still the Real Scorecard
Talkington’s remarks also highlighted something that matters even more to Microsoft investors than Copilot sentiment:
Azure growth. For all the headlines around AI assistants, the stock’s next move still depends heavily on cloud execution. Microsoft can endure product chatter if Azure keeps compounding at a strong rate.
That is why her focus on 38% to 39% year-over-year cloud growth is notable. It shows that investors are increasingly treating AI software as a narrative overlay and infrastructure as the actual earnings driver. If Azure comes in below expectations, the market can quickly ignore the rest.
Cloud Growth Versus AI Hype
The market has spent two years rewarding AI promises, but it is now demanding proof that the revenue math works. Azure is where that proof is most visible. It absorbs capital, powers workloads, and provides the operational foundation for Microsoft’s broader AI stack.
If Azure growth remains strong, then weaker Copilot sentiment may be survivable. If Azure slows, however, then Copilot criticism becomes part of a broader thesis that Microsoft’s AI premium is fading. That would be a more serious problem for the stock.
- Azure remains the main valuation anchor for Microsoft.
- AI assistant adoption affects narrative, but cloud growth affects earnings.
- Investors are now scrutinizing monetization, not just innovation.
- A slight miss in cloud growth could overshadow product announcements.
- Microsoft’s AI story is strongest when the infrastructure story is strong too.
What the Market Is Really Pricing
Microsoft’s valuation has compressed enough to invite bargain hunting, but that does not eliminate risk. The company is still priced as a premium compounder, which means execution must remain consistently strong. In that context, Copilot is less about a single product launch and more about whether Microsoft can sustain premium expectations across multiple businesses.
That is why the conversation around SaaS repricing matters. Microsoft may be a quality winner, but quality winners can still get rerated if growth cools or if their AI narrative looks less differentiated than expected.
SaaS Repricing and the Case for Buying the Dip
Several investors on the broadcast argued that Microsoft’s recent weakness may be creating an entry point. The logic is that the stock’s forward multiple has compressed materially, and that a company of Microsoft’s scale, cash generation, and ecosystem depth should benefit when the market separates durable software leaders from weaker peers.
That argument is plausible. Microsoft is one of the few companies with the distribution, capital, and enterprise relationships to turn AI into recurring revenue at scale. If the SaaS sector is being repriced, Microsoft could emerge as one of the survivors that deserves a premium.
Why Bulls Still Like Microsoft
Bulls are not just buying a story; they are buying a franchise. Microsoft has proven its ability to translate platform power into subscription economics, and that matters more in a tighter market. It is also diversified across productivity, cloud, security, developer tools, and gaming, which gives it resilience that smaller software firms lack.
The key is whether the AI transition can enhance that moat rather than muddy it. If Copilot becomes a habit-forming layer and Azure remains strong, the stock can recover even if the path is choppy. The market tends to reward platforms that keep winning while others are still trying to find product-market fit.
What Could Still Go Wrong
The downside case is not trivial. If Copilot remains a niche or clunky experience, and if Azure growth merely meets rather than beats expectations, investors could conclude that Microsoft’s AI premium was overextended. That would keep pressure on the multiple.
There is also the risk that Microsoft becomes too dependent on bundling to preserve growth. Bundles can sustain revenue, but they do not always create genuine preference. Over time, preference is what keeps enterprise software pricing power intact.
Apple, Energy, and the Rotation Trade
The broadcast also touched on a broader portfolio rotation theme, with Apple bought on weakness and Marathon Petroleum trimmed after a strong run. That matters because it shows investors are not making isolated AI calls; they are reallocating across sectors based on momentum, valuation, and macro expectations.
Apple’s case is instructive because it reflects a different kind of capital discipline. While Microsoft is spending heavily on AI infrastructure and product expansion, Apple is being viewed through a more measured lens. That contrast makes Microsoft’s capital intensity look more aggressive, even if strategically justified.
Apple as a Counterexample
Apple’s narrative in this discussion is less about AI hype and more about ecosystem durability. Some investors see it as a consumer services and hardware utility business with optionality around Siri improvement. That is a very different model from Microsoft’s enterprise AI push.
Talkington’s point that Apple’s capex is down year over year reinforces the idea that investors are paying attention to discipline as much as ambition. In a market obsessed with AI infrastructure, restraint can look like a feature. It is
not the same game Microsoft is playing, but it shapes relative preference.
- Apple is being viewed more as a disciplined ecosystem story.
- Microsoft is being judged on AI capex and enterprise monetization.
- Capital allocation discipline can influence valuation as much as product growth.
- Portfolio rotation is showing up across both tech and energy.
- The market is rewarding clarity in business models.
Energy’s Role in the Reallocation
The Marathon Petroleum trim reflects another classic rotation: taking profits from a strong cyclical trade and moving toward tech weakness. That is a reminder that capital is always chasing the next source of asymmetry. When energy cools and software sells off, the market can quickly change its leadership preferences.
For Microsoft, that creates both opportunity and pressure. It can attract capital from investors looking for durable growth, but only if its AI story looks cleaner than the rest of the market’s. If not, it risks getting lumped in with the broader software re-rating.
Strengths and Opportunities
Microsoft still has enormous strengths, and the Copilot backlash should not obscure them. The company controls one of the most important enterprise distribution networks on the planet, and it has the financial firepower to keep iterating. In the long run, that combination can win even if the first version of a product disappoints.
More importantly, Microsoft is not standing still. It is pushing deeper into agent governance, broader model support, and enterprise packaging in ways that could turn Copilot from a feature into a platform layer.
- Massive distribution through Microsoft 365, Windows, and Azure.
- Enterprise trust that many AI startups still have to earn.
- Strong cash flow to fund product iteration and infrastructure.
- Bundling power that can drive adoption even when standalone demand is soft.
- Platform integration that competitors may struggle to match.
- Agent governance opportunities through E7 and Agent 365.
- Potential upside if Copilot improves fast enough to become habitual.
Risks and Concerns
The biggest risk is that Microsoft’s AI strategy becomes too dependent on distribution and not enough on delight. If users keep seeing Copilot as a tolerated feature rather than a preferred assistant, engagement will lag and competitive differentiation will remain thin. That would be especially problematic if rivals keep improving faster.
There is also execution risk around packaging. The E7 strategy could be powerful, but it could also alienate customers if they view the suite as expensive bundling rather than value creation. In a cautious enterprise buying environment,
pricing power has limits.
- Weak user sentiment can undermine adoption even inside a strong ecosystem.
- Competitive pressure from Claude, ChatGPT, and others is intensifying.
- Bundle fatigue may make customers skeptical of premium suites.
- Azure concentration leaves the stock sensitive to cloud growth misses.
- AI capex pressure could keep margins under scrutiny.
- Perception risk rises when critics compare Copilot to disliked products like Teams.
- Monetization uncertainty remains if usage does not become habitual.
Looking Ahead
The next few weeks will matter because Microsoft has to prove that the criticism is not becoming a narrative. The April 29 earnings report will be watched for Azure growth, while the May 1 E7 launch will be scrutinized for whether it feels like a genuine product reset or just a pricing exercise. Those are different tests, but they are linked.
If Microsoft can show that Copilot usage is deepening, that Azure remains healthy, and that the new suite is a convincing response to enterprise AI demand, then the recent backlash may fade quickly. If not, the market may begin treating Copilot as an expensive distribution channel for a product customers still do not love.
Key things to watch
- Azure revenue growth and whether it meets the high-end investor range.
- Copilot engagement trends rather than just license counts.
- E7 uptake among large enterprise customers.
- Any product redesigns that simplify Copilot’s user experience.
- Competitive displacement from Claude, ChatGPT, or Google’s enterprise AI stack.
- Commentary from CIOs and IT buyers about workflow fit and governance.
Microsoft’s long-term position is still formidable, but the Copilot debate is a reminder that software empires are not sustained by reach alone. They survive when the product feels essential, the pricing feels justified, and the experience feels better than the alternative. If Microsoft can get those three things to line up again, Copilot may yet become the AI layer it wants to be; if not, the market will keep rewarding rivals that feel more like tools and less like obligations.
Source: Sahm
Microsoft's Copilot 'Feels Like Teams:' Why This Fund Manager Is Swapping It For Claude