INFORMS Analytics Framework

Analytics doesn't begin with data.
It begins with a question.

And it doesn't end with a model. The INFORMS Analytics Framework™ is a comprehensive structure for applying analytics and AI across the entire life cycle of an initiative, from initial business concept to sustained, value-delivering operations.

The IAF is the foundation for all three tiers of the CAP® certification, and the structure by which successful analytics efforts are approached and maintained.

Where would you like to start?

Explore the Framework  Learn About CAP Certification

80%

of AI projects fail

Fortune, 2022
RAND, 2024

The analytics-to-action loop is broken, and increased adoption of AI will only amplify existing flaws if structure and discipline are not put around it.

RAND researchers found that, by some estimates, more than 80% of AI projects fail to reach meaningful production deployment. They documented the root causes: wrong problem framing, poor data practice, technology prioritized over the problem, inadequate deployment, and no lifecycle management. The IAF provides repeatable practices and methods to avoid every one of them.

Why it matters

Seven reasons AI and analytics initiatives fail to deliver value

These are not technology failures. They are structural and human failures, the same ones RAND researchers found at the root of most AI project collapses. Better tools won't fix them. A better approach to applying analytics will.

  • Fragmented data
    Data is siloed across teams, systems, and functions, making it difficult to build the reliable foundation that analytics initiatives require.
  • Low analytics fluency
    Many stakeholders lack the skills to frame problems analytically, interpret outputs, or challenge model assumptions, even when good data and good models exist.
  • Siloed expertise
    Knowledge stays trapped in individuals or functions instead of being captured, shared, and reused across the organization.
  • Inconsistent processes
    Different teams apply analytics using different methods, criteria, and assumptions, with no shared structure to ensure consistency or reproducibility.
  • Skipped problem framing
    Without disciplined problem framing, AI and analytics efforts address the wrong question, producing technically correct outputs that solve the wrong challenge for the organization.
  • Execution and adoption gaps
    Even when models are technically sound, organizations fail to deploy them effectively or sustain their value over time, limiting impact and eroding confidence in analytics.
  • Stakeholder misalignment
    Without structured communication checkpoints, analytics teams and the stakeholders who depend on their work drift out of alignment, producing solutions that no longer reflect organizational needs by the time they are delivered.
The INFORMS Analytics Framework was built to address every one of these gaps.
What the IAF is

Not another framework. A comprehensive structure that sits above them all.

The IAF is not competing with CRISP-DM, Six Sigma, or MLOps. It provides the organizational context, shared language, and structured approach within which any analytics or AI methodology can be selected, combined, and applied to help organizations make better decisions.

"

Where other frameworks ask "how do we build this model," the IAF asks "how can this organization leverage analytics and AI to make better decisions, and how do we sustain that capability over time."

"

Existing frameworks like CRISP-DM, MLOps, and Six Sigma operate at the project execution level. They are valuable tools. But they don't provide a comprehensive structure for the entire analytics lifecycle. They don't address what happens before the model, or what happens after it's deployed. They don't build the shared language, repeatable practices, and disciplined problem framing that organizations need to ensure analytics delivers genuine business value consistently.

The IAF does. It moves analytics from isolated model-building to a repeatable, end-to-end approach that connects every initiative back to its business objectives. Within each domain, the IAF creates structured checkpoints for communication and agreement, ensuring that stakeholders, analysts, and leaders are aligned before work proceeds to the next stage. That discipline, integrating regular communication, problem framing, and organizational alignment at every stage, is what the IAF provides that no competing framework does.

Who it's for

Whether you build analytics, lead it, teach it, or depend on it, you belong here.

Analytics professionals

You build things that don't always get used. The IAF gives you the language and the community to change that.

Whether your title is data scientist, operations research analyst, management consultant, financial analyst, industrial engineer, or something the job boards haven't named yet, the IAF covers the full scope of what you actually do – not just the model-building part. It also gives you a structured way to keep stakeholders aligned throughout the work, so what you build reflects what the organization actually needs.

If you frame problems before anyone touches the data, you belong here. If you need your work to survive deployment, you belong here.

The C-suite

You've invested in analytics. The IAF helps you know whether it's working and why.

The IAF provides a structured approach that reduces risk, increases confidence, and builds the organizational capability to apply analytics and AI to make better decisions consistently. It gives you a shared standard for evaluating analytics work across every stage of the lifecycle, not just the outputs – and a structured way to ensure your analytics teams are not operating in isolation from the people whose decisions depend on their work.

If you're accountable for whether analytics delivers value, you belong here.

Educators and academics

You're training the next generation of analytics professionals. The IAF gives them a framework that works in the real world from day one.

Students who graduate with exposure to the full analytics lifecycle – not just the technical methods – enter the workforce ready to frame problems, earn stakeholder trust, deploy solutions, and sustain their value over time. The IAF provides a common structure that translates directly from the classroom to every industry and every job title your students will hold.

If you're preparing students for analytics work in industry, government, or consulting, you belong here.

Analytics leaders in transition

Moving from building models to leading teams? The IAF is built for this moment.

The transition from senior individual contributor to analytics manager is where the technical skills stop being the limiting factor. Framing, stakeholder alignment, deployment, and lifecycle management become the job. The IAF structures exactly that work – and the INFORMS community includes people who have navigated it.

If you're responsible for what happens after the model is built, you belong here.

"

“After years in analytics, I've watched countless projects succeed or struggle based on seemingly intangible factors. The INFORMS Analytics Framework brilliantly captures what many of us have learned through trial and error, and formalizes it into actionable guidance.”

"
Carlos A. Zetina, Decision Intelligence, FICO Xpress
How IAF is different

A fundamentally different kind of framework

Dimension Other frameworks INFORMS Analytics Framework
Orientation Project-based: analytics as a tool for generating insights Process-based: end-to-end structure for successfully applying analytics and AI
Scope Stop at models, insights, or data mining – a finite project with a clear start and end Extend to value realization and sustained impact through a continuous lifecycle with feedback and learning
Human judgment Assume decisions are purely objective Explicitly incorporate human judgment, expertise, and creativity alongside data and models
Decision-making Support one-off decisions, anchored to initial assumptions Enable dynamic, revisitable decision-making that evolves with changing conditions and new information
Organizational role Position analytics as a specialist function relying on top-down dissemination of insights Create a shared approach where the whole organization negotiates how to apply analytics and AI to make better decisions
Disciplinary scope Operate within a narrow analytics discipline Take a multidisciplinary approach integrating business, technical, and domain perspectives
Technology Tied to specific vendors or tools Vendor- and technology-neutral
Stakeholder alignment Treat analytics as a technical exercise, with communication handled informally or not at all Build structured communication checkpoints into every domain, ensuring stakeholders and analysts remain aligned throughout the lifecycle
Professional standard No unified professional standard Linked to CAP certification to codify and validate full-lifecycle capability
What the IAF stands for

Six attributes that define how the framework works

  • Integrated
    A shared foundation that gets everyone involved in an analytics initiative on the same page – across functions, levels, and disciplines.
  • Forward-looking
    Built for the age of AI and continuous change, not for a static project environment. The IAF is designed to evolve with the tools and conditions around it.
  • Outcome-driven
    Every domain in the IAF connects back to a decision and a measurable result. The framework does not treat model-building as the end goal.
  • Adaptive
    Supports revisable, dynamic decision-making that evolves with new information and changing conditions – not one-off choices anchored to initial assumptions.
  • Human-centered
    Explicitly incorporates human judgment, expertise, and creativity alongside data and models. Great decisions require more than great analysis.
  • Method-agnostic
    Vendor- and technology-neutral. The IAF works with CRISP-DM, Six Sigma, MLOps, or any other methodology – it provides the structure within which those tools operate.
The outcome

Increased confidence, reduced risk, and greater organizational understanding of how analytics and AI deliver value, and why these approaches hold up over time.

Want to explore first?

Not ready to dive in? Start with the details that matter most to you.