Across sessions at AlphaSummit 2026, our second annual user conference, the conversation revealed a major shift underway in how enterprises operate. As AI makes research faster and easier to access, the advantage now comes from how teams use it.
Leaders from across the business landscape — Salesforce, Pfizer, Mondelez, CAVA, Keurig Dr Pepper, sweetgreen, and others — discussed how they’re looking beyond internal metrics to outside signals, checking each signal against other data before acting, and turning confirmed insights into clear priorities. For example, Salesforce used broker research and expert interviews to judge an AI-for-service opportunity its internal data couldn't measure, then backed its recommendations with work on live deals.
At AlphaSummit, we also announced SuperAnalyst, our next-generation, always-on AI agent that executes high-value financial and strategic workflows. In this article, discover the top AlphaSummit takeaways for corporations and consulting firms, including how SuperAnalyst arms teams with intelligence they can defend, act on, and trust.
View full sessions from AlphaSummit 2026 here.
Uncovering Key Signals, Wherever They Surface
A corporate team’s default data sources can be genuinely useful. Metrics like win and close rates help answer questions that you already know to ask, such as, “How do our offerings stack up with peers?” But relying on these metrics alone does not solve for “unknown unknowns,” leaving potential blind spots.
Dan Hamilton, Salesforce’s SVP of corporate strategy, recounted identifying AI for service as a major growth opportunity in a market the firm already led during the “Are We Winning, Why, and What Should We Do About It” fireside chat at AlphaSummit. The problem: Salesforce’s internal data alone wasn’t enough to gauge the scale of the opportunity. With ever-more of the sales cycle happening informally outside of the RFP, external data is a major blind spot. Once Hamilton’s team added broker research, sell-side commentary, and expert interviews through AlphaSense, they could see what customers and other outside voices were saying about Salesforce in real time. This external research eventually helped shape the team’s understanding of Salesforce’s place in the market and who it needed to compete against.
We actually saw quickly that just looking at our own internal data around 'are we winning' was not enough. We had to look externally at some of these additional signals, and not just TAM data and market share, because that's such a lagging indicator. …I don't care what market share was 12 months ago. [I want to] know what it's going to be in the next year.
Cameron Vollmuth, head of investor relations at Acushnet Company, ran into a similar issue. Because his firm operates in a niche vertical with just a single publicly traded competitor, peer benchmarking alone would give a narrow read on overall market sentiment. Instead, Vollmuth starts his research with a specific question, which helps steer the search. For example, when assessing consumer health, the expert looks to the banks, which offer robust consumer spending data and also tend to report earnings before other sectors.
Anticipating Questions Before They Arise
The hardest questions your team receives on an earnings call or in a strategy review often surface first elsewhere, whether in a peer’s own results or a competitor’s new move. Monitoring those sources gives teams plenty of time to prepare.
For Vollmuth, tracking analysts who cover his company is a valuable source of question prep. For example, when peer Dick’s Sporting Goods reported weakness in apparel and footwear, he used it to prepare data-backed responses for Acushnet Company’s leaders ahead of his own call.
At the “How AI Is Rewiring Life Sciences Workflows” panel at AlphaSummit, Christopher Salatto, Pfizer’s head of competitive intelligence, outlined how he uses a similar tactic with drug pipelines. His pipeline teams track competitors’ assets, trial designs, inclusion and exclusion criteria, and mechanisms of action to inform how Pfizer develops and positions its own assets.
It's not only understanding what the assets are, it's also understanding the organizations that are bringing those assets to market. What are the things that are influencing those organizations? [We are] trying to figure out how those organizations are making decisions and how they may decide to progress the particular molecule that we're thinking about, what they might progress in tandem to that molecule, or how they may be positioning that molecule in the market.
Separating Genuine Insights From the Noise
A single data point can reflect a pattern or an isolated event, and signals from different sources don’t always agree. To solve this, AlphaSummit speakers described testing a signal against other data before acting on it, both to filter out one-off occurrences and to understand why their sources diverge.
During the “Before the Street Asks: How IR Teams Turn Market Signals Into Management Conviction” panel at AlphaSummit, Rishabh Bhansali, Mondelez’s director of investor relations, recounted a story of reconciling conflicting data from different sources. Bank data was showing robust consumer spending, while consumer staples companies were reporting category weakness and flat volumes. Bhansali mentioned that, because the same consumer can behave differently in different settings, both things could have been true at the same time. To find where the weakness is actually showing up, he combined bank data, retailer commentary, adjacent company signals, and Mondelez’s own internal data before deciding how the company should respond.
Meanwhile, during the “Reading the Business Like an Investor” fireside chat, Cindy Olsen, sweetgreen’s chief strategy officer, shared two anecdotes describing how she uses targeted signals to drive business decisions.
In one instance, a simple conversation with an old colleague who had no idea sweetgreen served hot plates prompted Olsen to rethink customer acquisition. The interaction signaled that sweetgreen could be doing much more to attract new customers, even as internal metrics showed that existing guests were visiting more often and digital growth was strong. Olsen then pivoted the marketing budget toward top-of-funnel awareness that prioritized outdoor advertising and local community engagement. In another example, she received a complaint from an investor that pickup times had grown from 20 minutes to 45 minutes. After pressure-testing the complaint against internal operational data such as on-time performance and complaint frequency, her team discovered the issue was widespread and implemented dynamic, restaurant-specific fixes.
I love those types of anecdotes because it allows me to go in, and we have all the data, where I can kind of see: Is this a bigger issue or not? …It was really understanding all the data, triangulating it, making sure it's a big enough opportunity that we should go after today.
Turning Validated Insights Into Organizational Action
Identifying a genuine insight does not tell you what you should do about it. The work lies in deciding where to focus, what needs to be done, and in what order. For example, sweetgreen’s Cindy Olsen described leveraging analyst research and expert insights in AlphaSense to help size the market opportunity in different category segments and meal times. Previously, evaluating menu items was done case by case, but these insights helped the team move to a more holistic way of targeting the categories where sweetgreen had a strategic “right to win.”
Chethan Mallela, Keurig Dr Pepper’s head of investor relations, described guiding investors through KDP's acquisition of JDE Peet's and its planned separation into two companies. He started by examining disclosures from previous corporate separations through AlphaSense, including disclosure timing and the surrounding investor messaging. The team then used this broader perspective to build their own deal strategy: designing the deal messaging around clear objectives, increasing disclosure step by step, and determining more granular specifics based on investor feedback.
Salesforce’s Dan Hamilton compares competitive intelligence teams to doctors who diagnose, prescribe, and follow up. His own decision framework weighs five factors behind every won or lost deal, drilling down on the two or three that matter most. He found that, in the voice agent space, product quality, human-like performance, and resolution-based pricing “kept coming up time and time and time again.” But taking that diagnosis and turning it into action requires both an “air game” of strategic initiatives and a “ground game” of live deals, workshops, and enablement assets.
We weren't just saying, 'Here's what the company needs to do to change.' We were also saying, 'And here's all the things that we are doing.' We're swarming all the open competitive deals. We're [finding] out all this feedback from them. We're writing X, Y, and Z. We're doing these workshops. And because we were also driving action and not just driving opinions, that's what helped the opinions that we had get even more traction than they would have otherwise.
Reinvesting the Time AI Gives Back
AI is helping enterprises expedite research workflows, but the scale of the benefit depends on how teams use the time savings. For AlphaSummit speakers, AI gains are showing up in the form of stronger, more defensible recommendations and allowing teams more time to spend on high-value work.
Chintan Oza, Merck’s executive director of enterprise strategy, said that, while AI has not dramatically shortened his team’s timelines, it has markedly improved overall decision quality. With AI handling research and data gathering that once required consultants, his own team has more time to devote to crafting high-quality recommendations to executives on the first pass, cutting out rounds of iteration.
Tricia Tolivar, CAVA’s chief financial officer, described taking a similar approach to restaurant operations during the “How CFOs Make Decisions in the Age of AI” panel. For the past 18 to 24 months, CAVA has been building its own internal data platform that now supports predictive prep, scheduling, and ordering, all part of an effort to remove “sand from the gears.”
What we're doing is continually showing examples and proof points to others on how it works and what it's allowed our team members to develop and explore that's not report generation or data analysis. It's really [about] partnering — sitting side by side and advising the business on where they can make those big bets and smart decisions or adjustments to drive improved returns.
Raising the Bar on Trust as AI Proliferates
General-purpose AI tools have made answers easy to find, and quality harder to control. AlphaSummit speakers described having checks in place to avoid the pitfalls of blindly relying on AI-generated outputs.
According to Pfizer’s Christopher Salatto, colleagues across the company already query general-purpose AI models directly, raising concerns about accuracy. His multifaceted response: to educate users up front while ensuring high-quality data on the backend. He strives to enforce a single source of truth and reduce silos across functions. In fact, all three IR panelists described avoiding drafting core messaging solely with AI, opting to verify critical outputs manually.
When colleagues get access to [LLMs], they start querying. And they don't always have a very keen eye as to whether the content they're getting back is accurate and precise enough for the question that they're trying to answer. So that's one of the jobs that I'm taking on within the organization: to educate, but also to find ways to make sure that the data that's flowing into those LLMs is of the highest quality, so that what's coming out is something that we can trust as an organization.
Building Intelligence That Agents Can Rely On
As AI agents take on more enterprise work, risks are compounding across functions. In a fragmented research stack, context gets lost between sources, tools, and workflows. That’s why it’s important to have a decision-grade solution that’s fully traceable and auditable.
Enter AlphaSense's SuperAnalyst, an always-on AI agent that plans, monitors, executes, and delivers entire workflows on your behalf. SuperAnalyst is capable of creating client deliverables, interactive dashboards, financial models, memos, presentations, and other finished outputs — and keeping that material constantly updated as situations evolve. Additionally, because SuperAnalyst combines your firm’s internal knowledge with AlphaSense’s industry-leading library of premium, proprietary content, every finding is backed by a trusted source.
Missed this year’s AlphaSummit? Catch up on five big things you missed, and reserve your spot at AlphaSummit 2027 here.
