As a Product Manager, how often have you given 100% to delivering a feature, only to see it rolled back later due to low adoption or minimal impact on your North Star metric? Quite often, right? In his book Evidence Guided, Itamar Gilad introduces the GIST framework, a practical approach to navigating the uncertainties of product and business development. It helps reduce feature failure rates and continuously validate assumptions to ensure real value is delivered to users.
#1 Goal
The first step is to clearly define a goal. A well-defined goal should be SMART—Specific, Measurable, Achievable, Relevant, and Time-bound, as well as both outcome-oriented and business-driven. When teams focus on outcomes rather than outputs, they gain the flexibility to explore different ways of achieving those outcomes. This encourages innovation and helps avoid the trap of building features that fail to make a meaningful impact.
Organizational leaders should start by setting Objectives and Key Results (OKRs) to define high-level objectives and the measurable results expected from them. From these OKRs, the company can identify its North Star Metric, the single most important metric that reflects long-term success, along with other top-level business KPI that everyone in the organization can align with.
To ensure alignment across the organization, leaders should also develop a metric tree. This breaks down overarching goals into actionable, team-level metrics that directly influence those goals. As a result, individual teams can define their own North Star Metrics that align with the company’s OKRs.
Even the roadmap should be framed in terms of outcomes, not just outputs. For example, instead of stating “Launch a chatbot,” a GIST style goal would be “Increase customer support satisfaction from 80% to 90%.” This shift keeps the focus on the value delivered, rather than merely on the feature shipped. Here's another example:
#2 Ideas
Once you’ve defined what needs to be achieved, the next step is to brainstorm ideas. Ideas can come from anyone—engineering, design, business, finance, and other stakeholders. It’s recommended to create a goal-level idea bank where all team members and stakeholders can contribute their suggestions.
From there, it's the responsibility of the Product Lead to evaluate and prioritize these ideas using the ICE framework—Impact, Confidence, and Ease. Each idea receives an ICE score, calculated as:
- ICE Score = Impact × Confidence × Ease
- Impact (max 10): Input should be gathered from team members and business stakeholders, focusing on how much the idea is expected to improve the North Star Metric (NSM) and relevant business KPIs. This can be supported by insights from past experiences with similar features.
- Ease (max 10): This should be evaluated in collaboration with engineering stakeholders, based on technical feasibility and effort.
- Confidence: This should be grounded in evidence, such as user research, product usage data, or past experiments. To assess this, Product Leads can refer to a Confidence Meter, which helps quantify how strongly the available data supports an idea.
This process ensures that ideas are not just creative, but also actionable, evidence-backed, and aligned with the overall goal.
#3 Steps
Next, you should conduct a series of Build-Measure-Learn feedback loops. Each loop will provide valuable input to update your Confidence Meter and help you refine the ICE scores of the prioritized ideas. These loops should be completed within one week, focusing on small, targeted initiatives that test assumptions and gather evidence. Keeping each step small promotes rapid iteration, making it easier to learn and adapt in the next cycle.
These steps can take various forms, such as: user research, A/B experiments, limited-time referral campaigns, usability testing of interactive prototypes, etc.
To structure and guide these activities, the AFTER framework can be used:
- Assessments – e.g., expert reviews, heuristic evaluations
- Fact-finding – e.g., market research, user interviews
- Tests – e.g., smoke tests, early-stage prototypes
- Experiments – e.g., A/B tests, minimum viable products (MVPs)
- Release results – e.g., live product usage data
#4 Tasks
Tasks represent the final execution step in the GIST framework. This is where product teams move into delivery, typically using agile methodologies. Tasks should only be created for ideas that have gone through the Steps process and have sufficiently high revised confidence scores. This selective approach reduces waste and increases the likelihood that engineering efforts lead to meaningful impact.
To maintain transparency across all stakeholders during execution, product teams can use a GIST Board to track progress. This visual tool helps align everyone on what’s being worked on, what’s next, and how each task ties back to validated ideas and strategic goals.
To see real-world examples of how the GIST framework has improved the success rate of product releases across various companies, read Evidence Guided by Itamar Gilad.