Why Most Analytics Teams Won't Survive the Next 5 Years

Summary

The speaker argues that analytics teams are facing a fundamental change: stakeholders increasingly believe they can get data answers directly from AI tools or other, cheaper alternatives. Even when those alternatives produce unreliable results, that perception puts pressure on every step in a data team’s work. Teams therefore need to demonstrate the value they add—not merely process requests or translate requirements.

Experiences illustrating the problem

• Pontis: The speaker joined a company whose marketing teams depended on R&D database administrators to build customer segments and report campaign results. The administrators saw this as outside their role, leaving marketing without timely support. By noticing the gap, learning the marketing team’s needs, and taking on the work, the speaker helped create and grow a centralized analytics team. Because that team worked across projects, it also gained a broad view of how the company operated and was later asked to help systematize its marketing approach.
• PayPal: In risk management, the speaker encountered a long chain of handoffs between risk analysts, product managers, subject-matter experts, business systems analysts, BI and ETL developers, and data modelers. The process made even small changes take weeks or months. The speaker’s point: every step in a value chain has a cost and must justify its contribution. If a team only relays requirements, it is vulnerable to being bypassed.

What analytics professionals contribute

The speaker says effective analytics requires four things:

1. The right data must be available.
2. The right questions must be asked; good data cannot rescue a badly framed question.
3. The right answers must be obtained and checked.
4. The right conclusions must be drawn from those answers.

The distinction between an answer and a conclusion is illustrated with survivorship bias. During World War II, Abraham Wald reportedly advised examining the bullet holes on aircraft that returned. The missing damage locations mattered most: planes hit there may not have returned. The lesson is that even correct observations can lead to the wrong action if their context and limitations are misunderstood.

The speaker sees the core role of data professionals as creating clarity in a noisy environment: finding reliable data, framing useful questions, validating results, interpreting them, and helping decision-makers understand what matters and what to do next.

Capabilities and practices to develop

The speaker recommends that data professionals:

• Build business context. Understand how the business works, who its stakeholders are, what they are trying to accomplish, and which problems are priorities. Data work should not be disconnected from the business it supports.
• Build direct relationships with users and stakeholders. Do not rely entirely on intermediaries to translate requirements. If work arrives through Jira tickets, identify the ultimate stakeholder and speak with them directly.
• Use curiosity and critical thinking. Ask what problem a request is meant to solve, whether the proposed work will solve it, and whether there is a better approach. Be willing to challenge assumptions and identify when a requested solution will not work.
• Take initiative and accept professional risk. Go beyond waiting for assigned tasks: ask operators about their challenges, ask leaders about their major problems, and question why a product or feature is being built.
• Invest in networking. The speaker argues that relationships and access to people who know and trust you are increasingly important as technical execution becomes easier to obtain elsewhere.

Overall message

As AI tools and other alternatives raise expectations—and sometimes generate misleading results—data teams may face pressure to justify their role. The speaker does not claim that these tools reliably replace analytics professionals; rather, stakeholders may believe they can. Professionals who only fulfill requests or pass requirements along are at risk. Those who combine technical knowledge with business understanding, sound judgment, direct relationships, initiative, and the ability to create clarity have a stronger future.

Speakers and sources featured

• Speaker: Shahar/Shachar Meir, a data executive and consultant; the subtitles render his name inconsistently.
• People and examples cited by the speaker: An unnamed Pontis CEO; an unnamed fintech senior director; an unnamed vice president who used Claude to query a data warehouse; PayPal risk analysts and associated teams; Pontis marketing and R&D teams; and Abraham Wald, cited in the World War II aircraft example.
• Other organizations mentioned: Pontis, PayPal, Meta, Facebook, Instagram, and Messenger. No other live speakers are identifiable in the subtitles.