I asked Analytics8’s seven practice leads to share what they believe will matter most in 2026 across data analytics, data strategy, AI, engineering, platforms, delivery, and operations. Below you’ll find what they shared. These are not theories or trends; these are priorities shaped by helping our clients solve their modern data challenges with both proven and emerging solutions. Below you’ll find: Where Our Experts Agree (and disagree) on 2026 Priorities ↵ Priority #1: Fix Analytics Adoption Issues with Standardized Definitions ↵ Priority #2: Use Governance to Make AI Explainable and Controllable ↵ Priority #3: Get Data Engineering Fundamentals Right Before Scaling AI ↵ Priority #4: Align Data Teams with Business Outcomes, Not Delivery ↵ Priority #5: Build Only the Platforms Your Organization Can Govern and Adopt ↵ Priority #6: Make AI Useful with Iterative Progress ↵ Priority #7: Strengthen Project-Planning Discipline Amid AI-Accelerated Delivery ↵ But first… Where our experts agree (and disagree) I chatted with seven data analytics leaders at our organization to get a pulse check on what they believe data leaders must focus on this year. These data analytics experts lead practice areas in our organization — roles earned through demonstrated client results, deep technical expertise, and proven ability to solve complex problems at scale. Combined, they bring more than 100 years of experience across the full data and analytics lifecycle, from strategy and governance to engineering, platforms, AI, and delivery. While their answers reflected a wide range of data and analytics focus areas, one theme surfaced across every discussion: organizational adoption is essential for 2026 success. But getting there isn’t easy. In this clip, our 7 practice leads explain why data and analytics success depends most on organizational adoption. “No matter how successful modernization is, how much the execs buy in, how innovative your design is, if the organization isn’t adopting the assets being built, it doesn’t matter how good they are.” – Travis LaMont, Analytics8 Project Management Practice Lead Where perspectives diverged was in where they believe organizations should prioritize their AI investments in 2026. Analytics8’s seven practice leads split on where to prioritize AI investments, with more than half choosing Model Context Protocol as the foundational starting point. When asked to choose between MCP and Agentic Frameworks, MCP edged out as a narrow winner. However, our experts agreed the answer ultimately depends on an organization’s analytics maturity. Our AI Practice Lead John Bemenderfer framed it clearly: “It’s not either/or. MCP is about connectivity. Agentic is about orchestration. MCP is foundational, then you build agentic on top of that.” Our Analytics Practice Lead Marty Lyman added the caution: “Agentic frameworks are only as good as the data foundation that makes them up. No matter what, there has to be a human element checking that you’re achieving real ROI.” Perhaps Data Strategy Lead Christina Salmi’s five-year-old son said it best. After listening to her explain MCP and AI, he asked: “But mommy, why is that important?” She laughed. “That’s such a smart question. Adults are not asking that.” The answer to “why is it important” depends on what problem you’re solving. The priorities below keep you anchored to that question, so that what gets built matters to the business and is adopted across the organization. What Data Leaders Need to Get Right in 2026 Marty Lyman, Analytics Practice Lead Marty is a Principal Consultant at Analytics8 with nearly seven years at the firm, specializing in analytics development. He’s worked with clients including Anheuser-Busch InBev, Crocs, and AbbVie, managing technical delivery and pushing best practices across business intelligence tools. When he’s not building analytics solutions, he’s working on his golf game or following Chicago sports. 2026 Priority: Fix Analytics Adoption by Standardizing What the Numbers Mean The biggest risk in analytics development is not lack of tools. It is lack of trust. Teams have dashboards, self-service platforms, and now AI-driven interfaces, yet adoption still breaks down because people do not believe they are looking at the same truth. Analytics only works when the organization agrees on what success looks like. Why Analytics Still Fails in Real Organizations Most analytics tools deliver exactly what they promise. The problem shows up after deployment. Different teams define the same KPIs differently. Revenue, margin, and performance metrics drift across departments. Leaders see conflicting numbers and lose confidence. Business users stop relying on shared dashboards and export data to rebuild logic themselves. This is where adoption dies. Not because analytics is unavailable, but because it is inconsistent. How AI Amplifies the Issue AI-powered analytics and natural language querying make this problem more visible. These tools answer questions faster, but they do not resolve disagreement about definitions. They amplify whatever logic already exists. Without standardized KPIs, governed semantic layers, and shared calculation logic, AI accelerates confusion instead of insight. What to Focus on in 2026 Analytics leaders should shift effort away from building more outputs and toward fixing the foundations those outputs depend on. That starts by identifying where critical KPIs are defined differently across teams, then consolidating that logic into shared, governed semantic layers. AI and MCP-style integration frameworks can help surface redundancies, conflicting models, and unused logic that quietly undermine trust. Once meaning is standardized, self-service and AI finally work the way they were promised to. Bottom Line: The future of analytics is faster and more automated. But speed does not matter if the organization is not aligned. In 2026, analytics development succeeds when meaning comes first. Christina Salmi, Data Strategy & Data Governance Practice Lead Christina is Analytics8’s Managing Director of Data Strategy, with more than 18 years in data and analytics consulting, working across the full data and analytics lifecycle. Over the last several years, she’s developed Analytics8’s data strategy and data governance practices. When she’s not helping organizations operationalize AI responsibly, she’s explaining MCP to her five-year-old son—who asks better questions than most adults. 2026 Priority: Use Governance to Make AI Explainable and Controllable In 2026, governance is no longer about policy or documentation. It is the control layer that makes AI usable at scale. As AI becomes embedded in analytics and decision-making, organizations need a way to understand, explain, and trust what those systems produce. Governance is what makes that possible. What to Focus on in 2026 Ontologies Ontologies move governance beyond static definitions. They capture relationships, context, and rules that describe how the business actually works. For most organizations, this means taking business knowledge that already exists and making it explicit. Instead of relying on undocumented assumptions, leaders should push teams to formalize definitions, relationships, and rules once, then reuse them across analytics and AI workflows. This is what turns AI from a black box into something the business can stand behind. Operationalizing AI Operationalizing AI requires intent and sequencing. Leaders should map AI use cases by effort and impact, start with low-risk enhancements that build confidence, and use those foundations to support more transformative initiatives over time. Just as importantly, data strategy and governance need to be revisited regularly as models improve and standards evolve. A static strategy will fail in a fast-moving AI environment. Modernizing Your Data Strategy The real challenge now is controlling how meaning and logic flow through AI-enabled systems. A modern data strategy should create structure without slowing progress. It should support experimentation while still defining clear boundaries for how data and AI are allowed to operate. Bottom Line: Governance is not a brake on innovation. It is what allows organizations to move quickly without losing control. In 2026, the teams that succeed with AI will be the ones that made meaning and responsibility non-negotiable. Simon Collis, Data Engineering Practice Lead Simon is a Principal Consultant at Analytics8 with nearly 10 years at the firm, specializing in data engineering, data architecture, and BI design. He’s built scalable data systems for clients including Anheuser-Busch InBev, GSK, Crocs, and Educause. When he’s not working, he’s on the tennis court — his USTA team has made PNW Sectionals three years running. 2026 Priority: Get the Fundamentals Right Before Scaling AI Data engineering success will not come from chasing the newest platforms or AI capabilities. It will come from doing the basics well, consistently, and with clear business intent. Strong foundations still matter. Without them, everything built on top, including AI, becomes fragile. Where Teams Get Distracted Many engineering teams fall into the trap of treating data work as a series of interesting technical projects. New tools, new platforms, and new AI initiatives move quickly, but they are often disconnected from measurable business impact. When teams cannot clearly answer why a project matters or how it affects margin, performance, or risk, adoption suffers and executive support fades. The issue is rarely effort. It is focus. Why AI Raises the Bar for Engineering Discipline Poor data quality is easy to spot in a dashboard. It is much harder to detect once it feeds AI systems. Bad data does not just produce bad answers. It degrades models over time in ways that are harder to trace and correct. This makes data quality, consistency, and auditability more critical than ever for engineering teams supporting AI workloads. AI does not compensate for weak foundations. It amplifies them. What Data Engineering Leaders Should Do in 2026 Engineering teams should double down on the fundamentals that have always worked. That means reliable ingestion pipelines, clear and efficient data models, and maintaining data at an atomic level so issues can be traced and corrected. Proven modeling approaches like dimensional data modeling are not outdated. They are what make data understandable to both humans and machines. At the same time, engineering work needs to be tied directly to business outcomes. Teams should be able to explain not just how a system works, but why it exists and what value it delivers. Where MCP Fits In MCP opens the door to practical acceleration, especially in modernization efforts. One promising use case is using MCP-connected systems to help translate and refactor legacy code into modern frameworks more efficiently. This kind of work saves time and cost, but it only succeeds when the underlying data structures are clean and well understood. MCP is an accelerant, not a substitute for good engineering. Bottom Line: AI changes what is possible, but it does not change what is required. In 2026, the engineering teams that succeed will be the ones that resist shortcuts, invest in strong foundations, and stay grounded in business value. Hart Shuford, Data Team as a Service Practice Lead Hart is a Consulting Director at Analytics8 with nearly four years at the firm and 13 years in data and analytics. She oversees a complex portfolio of data analytics engagements for clients including Bristol Myers Squibb, GSK, AbbVie, and Driven Brands, focusing on smooth delivery and strong client outcomes. Before moving into data consulting, she was an infectious disease epidemiologist. 2026 Priority: Align Data Teams to Business Decisions, Not Just Delivery In 2026, the success of data teams will hinge on how closely their work aligns to real business outcomes. Speed and technical execution matter, but they are not enough on their own. Data teams deliver the most value when they understand the decisions they are supporting and shape their work around those decisions from the start. Where Data Teams Miss the Mark Many organizations still treat data teams as delivery engines. The focus is on throughput, capacity, and getting work done faster. While this solves short-term needs, it rarely changes how the business operates. When data teams are disconnected from decision-making, insights stop at analysis. Reports get delivered, but concrete next steps are not clear to the business. Analytics only offers hindsight instead of something they use to direct action and drive outcomes. Why AI Changes Expectations AI makes analysis easier and more accessible, but analysis alone does not drive outcomes. The real advantage comes when insights are embedded directly into workflows with the right guardrails in place. As AI expands what data teams can produce, the ability to operationalize insights responsibly becomes essential. Teams that can close the gap between insight and action will outperform those that stop at reporting. What to Focus on in 2026 Data leaders should expect their teams, whether internal or delivered through a service model, to be accountable to outcomes, not just outputs. That means aligning work to specific decisions, metrics, and operational goals. It also means empowering data teams to challenge assumptions, shape solutions, and design analytics that are durable and scalable rather than one-off deliverables. In a Data Team as a Service model, the highest value comes when teams are trusted partners in the business, not supplemental capacity. Bottom Line: In 2026, data teams will be judged by how they drive outcomes for the business, not technical throughput. The organizations that succeed will be the ones that embed data and analytics into daily decision-making and empower teams to drive business outcomes in real time and de-emphasize passive or generalized reporting. Ed Pearson, Platform Modernization Practice Lead Ed is a Principal Consultant at Analytics8 with 20 years in data and analytics. He architects data estates for clients including Databricks, Coca-Cola, and Yum Brands, specializing in Azure, AWS, and Snowflake. Ed founded DataTune, a data conference for 500+ professionals in Nashville, and regularly speaks at industry conferences. When he’s not building data platforms, he’s building Legos with his granddaughter. 2026 Priority: Build Only the Platforms Your Organization Can Govern and Adopt A modern data platform is not defined by how advanced the technology is. It is defined by whether the organization can govern it, trust it, and use it as AI scales. Many platforms fail not because they lack capability, but because they were never built to support adoption at enterprise scale. What Has Not Changed The fundamentals of a strong data platform are still the fundamentals. Reliable ingestion. Well-modeled data. Scalable infrastructure. Those requirements have not gone away. What has changed is that gaps in data quality, ownership, and structure have become impossible to ignore. Governance is Now Non-Negotiable For years, governance was treated as an add-on. Platforms were built first. Metadata, access controls, and stewardship were addressed later, if at all. That approach must be flipped in an AI-driven environment. If machines are going to operate on enterprise data, they need context. They need to know what the data represents, how it can be used, and who is accountable for it. Governance is no longer a supporting function. It is a core platform capability. Where Automation Fits in Data Platform and Where It Doesn’t Agentic and automated data engineering capabilities show where platforms are headed. They can accelerate development, reduce manual effort, and free up time for strategic decision-making. But automation only works when the platform underneath it is clean and governed. Otherwise, it simply moves bad patterns faster. Bottom Line: In 2026, platform modernization is about readiness, not novelty. The organizations that succeed will be the ones that build platforms designed for governance and adoption first, so AI can scale without introducing risk or chaos. John Bemenderfer, MCP & AI Workloads Practice Lead John is a Principal Consultant at Analytics8, entering his fifth year with the firm and 11 years in data and analytics. He specializes in generative AI, agentic frameworks, and MCP-based systems, working with clients including a large Southern U.S. convenience store chain and the country’s leading beauty retailer. John publishes regularly on Databricks, MCP, and generative AI. Fun fact: he has an identical twin brother. 2026 Priority: Make AI Useful with Iterative Progress In 2026, the real challenge with AI is not getting answers. It is getting AI to do something useful with those answers. Most organizations are still stuck in a pattern where AI can explain, summarize, or recommend, but not execute. There is value in summarization and explanation, but what people really want is something that helps them do the work. When AI systems can’t do that, adoption stalls, and so does ROI. Why AI Has Historically Stopped Short Until recently, AI systems were hard to connect to real enterprise environments in a consistent way. Every new model provider, tool, or workflow required custom integration. Want to swap from Anthropic to OpenAI or introduce an open-source model? You’re rewriting code. That made automation brittle, slow, and risky. As a result, teams kept AI safely boxed into analysis and experimentation. It could inform decisions, but not participate in them. What Changes the Equation Model Context Protocol (MCP) removes much of that friction. By standardizing how AI systems connect to data, tools, and applications, MCP makes it possible to build execution paths once and reuse them across model providers and interfaces. The same tooling works for interactive conversations or agentic workflows. This is what allows organizations to iterate quickly, improving outputs in days instead of months, which drives adoption and keeps pace with how fast this space is moving. Where Leaders Need to Be Careful Execution raises the stakes. Once AI is allowed to act, questions of access, guardrails, auditability, and recovery matter immediately. Leaders need to be clear about where AI is allowed to operate, what decisions it can influence, and how issues are detected and corrected. Start with clearly defined, targeted use cases, not expansive ones. Incremental gains in aggregate can have a sizable impact. The goal is not full autonomy. It is reliable, bounded execution. Bottom Line: AI value will not come from better summarizations or explanations. It will come from taking action based on how your organization operates, within your own systems, using your own data and processes. The organizations that see real ROI will be the ones that bridge the gap between chatting and doing, increase adoption within their organization, and iterate as quickly as the AI industry is evolving. Travis LaMont, Project Management Practice Lead Travis is a Program Director at Analytics8 with over 10 years at the firm and 24 years in data and analytics. He’s led large-scale data and BI programs for clients including Driven Brands, GSK, and AbbVie, specializing in platform modernization and delivery execution. When he’s not managing complex data initiatives, he’s at Soldier Field as a Chicago Bears season ticket holder. 2026 Priority: Strengthen Project-Planning Discipline Amid AI-Accelerated Delivery Modern platforms, strong data foundations, and AI capabilities only matter if the organization adopts what is built. Project management is where that adoption is won or lost. Where Analytics Initiatives Break Down Many analytics efforts fail quietly. The work gets delivered, but it does not change behavior. Dashboards go unused. Data products stall after launch. Teams move on to the next initiative without knowing whether the last one mattered. This happens when delivery loses sight of the problem it was meant to solve. Without clear outcomes, even high-quality analytics becomes shelfware. Why AI Raises the Bar for Delivery AI is compressing timelines and expanding what teams can build. That makes delivery discipline more important, not less. As projects move faster, the risk of misalignment increases. Without strong frameworks, teams build impressive solutions that drift away from the original business need. AI can accelerate delivery, but it cannot correct poor planning or unclear goals. What Leaders Should Focus on in 2026 Data and analytics leaders should prioritize building scalable, repeatable delivery frameworks. That means defining clear steps for how work moves from intake to execution to adoption. It means aligning every project to a specific business problem and being able to articulate how success will be measured. It also means keeping teams focused on outcomes throughout delivery, not just at kickoff. Strong project planning is central to this. Breaking complex initiatives into predictable, sequential work creates transparency for both teams and stakeholders. It sets expectations, reduces risk, and keeps delivery grounded in reality. Bottom Line: In 2026, the differentiator will not be what analytics teams can build. It will be what they can deliver, communicate, and make stick. The organizations that succeed will be the ones that treat project management as a strategic capability, not an afterthought. Talk With a Data Analytics Expert Key Takeaways Organizational adoption is the unifying challenge across all seven practice areas. No matter how advanced your technology, platforms, or AI capabilities, success hinges on whether your organization actually uses what you build. AI amplifies existing problems at scale. If your KPIs are inconsistent, data quality is weak, or governance is an afterthought, AI will accelerate those issues faster than you can fix them. Standardized definitions come before automation. Different teams defining the same metrics differently kills trust and adoption. Analytics leaders must consolidate KPIs into shared, governed semantic layers before AI can deliver on its promise. Governance shifts from documentation to control layer. It’s what makes AI explainable, controllable, and trustworthy at enterprise scale, not a compliance checkbox addressed after platforms are built. Data engineering fundamentals matter more than new platforms. Reliable ingestion, clear data models, and atomic-level data maintenance are what make AI workable. MCP accelerates modernization but only when underlying structures are clean. Data teams should align to business decisions, not delivery metrics. Teams measured by throughput miss the mark. Success means embedding analytics into decision-making workflows and proving measurable impact. AI value comes from execution, not explanation. Most organizations use AI to summarize and recommend. Real ROI comes when AI can take action within your systems with clear guardrails and auditability. Project management discipline becomes more critical as AI compresses timelines. Faster delivery doesn’t fix poor planning or misaligned goals. Strong frameworks keep work grounded in business outcomes. Bottom line for data leaders: The organizations that succeed in 2026 will standardize meaning, operationalize governance, ground work in business value, and build for adoption — not just technical capability.
