• Consider Zeta Global (ZETA) for its reported growth and cash-flow momentum, but treat the CEO’s claimed valuation discount as unverified and watch for measurable revenue from its enterprise AI expansion.
• Palantir (PLTR) offers exposure to enterprise data and AI, but weigh its growth and profitability against the high valuation discussed; no price target was provided.
• For Meta (META) and Alphabet (GOOGL/GOOG), track sustained AI usage and monetization rather than early download figures or product announcements alone.
• Approach high private-company AI valuations cautiously, and wait for confirmed IPO details before acting on OpenAI or Anthropic.

Zeta Global (ZETA)

• CEO David Steinberg said Zeta rebuilt its platform to make data and AI central to its product, beginning in 2017 and rolling it out in 2020. He credited that strategy with helping drive company growth.
• He cited company-reported performance of more than 30% annualized revenue growth, 50% EBITDA growth, and 75% free-cash-flow growth over the prior four years. He also said Zeta had beaten and raised guidance in 20 consecutive quarters.
• Zeta’s platform is used for customer acquisition, retention, and monetization. The company says its data and AI help clients identify consumer intent and improve marketing results; Steinberg cited a Forrester survey estimating $6–$7 in return for each $1 spent through the platform.
• The company is expanding beyond marketing into enterprise intelligence, including inventory, logistics, finance, fraud detection, and cybersecurity. Steinberg described Athena OS as a potential platform for these uses.
• Zeta announced partnerships with Palantir and OpenAI, among others, and said its data cloud is not exposed to external foundation models. It also announced a custom-chip initiative, with manufacturing and hosting partners expected to fund the hardware.

Takeaways

• Zeta’s investment case, as presented by its CEO, rests on continued growth, strong customer expansion, and successfully extending its platform beyond marketing.
• Track whether new enterprise products and partnerships produce measurable revenue and customer adoption; announcements alone do not establish commercial success.
• Steinberg said Zeta trades at a 70% valuation-multiple discount to its comparison group and expects that gap to narrow if execution continues. This is management’s view, not an independent valuation conclusion.
• The discussion did not provide a price target. The CEO also noted that investors have historically viewed marketing spending as vulnerable when business conditions weaken, though he said Zeta grew during COVID.

Palantir (PLTR)

• Steinberg described Palantir as Zeta’s partner in bringing Zeta’s data cloud onto Palantir’s Foundry platform. He said the arrangement could improve the speed of Zeta’s data-driven decisions.
• He praised Palantir’s commercial business and said Zeta’s relationship had also led to government opportunities, including Zeta’s addition to FedRAMP for government contracting.
• Palantir was cited as a high-performing comparison for software metrics. Steinberg said it was the only company in Zeta’s peer group with a higher net retention rate and Rule of 40-style performance.
• The host and Steinberg discussed Palantir’s very high valuation relative to its sales. Steinberg pointed to its cash-flow and profit growth in response, while the host emphasized the unusually large valuation multiple.

Takeaways

• The discussion highlights Palantir’s potential as both a software peer and a strategic partner for enterprise data applications.
• Investors can weigh its growth and profitability against its valuation; the conversation offered no price target or valuation conclusion.
• The partnership may benefit Zeta, but the transcript did not quantify its expected financial contribution.

Snowflake (SNOW), CrowdStrike (CRWD), and MongoDB (MDB)

• Steinberg named these companies as part of a group of software companies he considers comparable to Zeta.
• He discussed software metrics such as net revenue retention, growth, operating margin, and free cash flow, but did not give individual performance figures for Snowflake, CrowdStrike, or MongoDB.

Takeaways

• The transcript suggests comparing software companies on customer expansion, growth, margins, and cash generation—not just revenue growth.
• It does not provide enough company-specific detail to support a view on any of these stocks.

Meta Platforms (META)

• The speakers discussed Meta’s AI assistant, Muse, and its potential to support personalized agents and commerce. Steinberg said Meta’s scale and distribution could be meaningful advantages.
• He cited 10 million downloads in a week for the product, while noting that it was still early to assess adoption.
• Steinberg said Zeta partners with Meta and considers its AI efforts interesting. He also referred to Meta’s investment in a company that helped bring in the executive leading that effort.

Takeaways

• The investment theme is whether Meta can turn its large user base and AI products into useful consumer and commerce experiences.
• The speakers cautioned, in effect, that early download numbers do not yet prove sustained usage or monetization. No stock recommendation or price target was given.

Alphabet (GOOGL/GOOG) and Google

• Steinberg argued that distribution is a competitive advantage and pointed to Google’s search products and Chrome as examples.
• The host said Gemini’s integration across Google products, including Chrome, made it useful in everyday tasks. Steinberg agreed that Alphabet should not be underestimated.
• Both speakers suggested that different AI models may serve different tasks, rather than one model winning every use case.

Takeaways

• The discussion favors considering distribution and product integration alongside model quality when assessing AI companies.
• It did not provide a specific view on Alphabet’s valuation or a price target.

OpenAI (private)

• Steinberg described OpenAI as a Zeta enterprise partner and said Zeta’s Athena assistant is integrated with ChatGPT.
• He characterized OpenAI as strong in consumer AI and said its advertising business was growing faster than he had expected. These are his assessments, not independently verified figures.
• The speakers discussed a possible OpenAI public offering, with the host referring to a potential Q1 or early Q2 2027 timeline. This was presented as an expectation, not a confirmed date.

Takeaways

• OpenAI’s potential investment opportunity may eventually depend on its ability to turn consumer reach and enterprise partnerships into durable revenue.
• It is a private company in the discussion, so the transcript offers no directly investable stock or valuation analysis.

Anthropic (private)

• The speakers discussed Anthropic’s focus on coding and enterprise revenue, and compared its pace of growth with OpenAI. No specific revenue figures were provided.
• The host referred to an S-1 and anticipation of a possible public offering in the coming weeks; the transcript does not establish a confirmed IPO date.
• Steinberg said the performance of a potential Anthropic IPO could influence the broader market’s view of AI companies.

Takeaways

• Any public offering would give investors a more direct way to assess Anthropic’s growth, business model, and valuation, but the transcript does not confirm when that might happen.
• The speakers’ comments indicate that investor appetite for new AI listings could affect sentiment toward other AI-related stocks.

AI applications, agents, and enterprise software

• Steinberg described AI applications as a major opportunity, arguing that much of the value from the AI transition could accrue to the application layer.
• He distinguished between AI-native products that use proprietary data to generate business intelligence and workflow tools that primarily reduce costs.
• The speakers discussed AI agents, inference, and personalized commerce. Steinberg said smaller, task-specific models can be more appropriate for some business uses than expensive frontier models.
• He argued that open-source models can be useful but may require significant customization and labor. He expects both open and closed models to have a role.
• The speakers named Salesforce (CRM), Adobe (ADBE), Workday (WDAY), and ServiceNow (NOW) as software companies whose shares had fallen sharply during an earlier AI-related reassessment and later recovered. They used this as an example of how quickly market narratives can change.

Takeaways

• A useful theme for investors is to distinguish companies that can demonstrate measurable customer returns from those whose AI offerings are still mostly promises.
• Consider both the potential benefits of AI-driven efficiency and the possibility that new products could disrupt existing software businesses.
• The transcript provides no specific stock recommendations, price targets, or timeline for the broader AI adoption cycle.

AI infrastructure and private-company valuations

• Steinberg compared today’s AI investment cycle with the dot-com era. He argued that many public companies currently funding AI infrastructure are highly profitable and could service their debt from existing cash flows.
• He contrasted those companies with some private AI businesses raising large sums at high valuations, which he said may struggle to justify the capital raised and valuations assigned.
• He also suggested that improvements in efficiency and compression could change future infrastructure needs, as technical advances changed the economics of internet capacity after the dot-com boom.

Takeaways

• The discussion points to two separate questions for investors: whether AI demand will grow, and whether individual companies can earn adequate returns on the capital they are investing.
• Treat high private-company valuations and large AI funding rounds cautiously; Steinberg specifically raised concerns about some private businesses’ ability to justify them.
• The transcript did not identify a specific infrastructure stock as a recommendation or provide a price target.

Other companies mentioned

• Apple (AAPL), Walmart (WMT), and JPMorgan Chase (JPM) were cited as examples of companies that benefit from distribution or reach.
• IBM (IBM) and its acquisition of Red Hat were used to illustrate that open-source technology can require substantial customization and labor.
• Yahoo and Qwest Communications were discussed as historical dot-com-era examples. The speakers noted that technology advances changed the value and use of internet infrastructure over time.

Takeaways

• These companies were mentioned as examples or historical comparisons, not as specific investment recommendations.
• No cryptocurrency was mentioned in the transcript.