GIST Planning
Replace your rigid roadmap with a 4-level planning system. The GIST framework aligns goals, ideas, experiments and tasks to deliver what truly matters.
Description
The GIST framework is a product planning system created by Itamar Gilad, former Gmail growth lead at Google. It structures work into four hierarchical levels: Goals (measurable long-term objectives), Ideas (hypotheses for reaching those objectives), Step-projects (small experimental projects to test each idea), and Tasks (daily activities managed in sprints). Its purpose: replace traditional roadmaps, which are too rigid and often disconnected from ground reality, with an adaptive planning framework. A classic roadmap works as a waterfall: you define a plan for 6 to 12 months, then execute. The problem is that internal data from Google and Bing show that only 10% to 20% of experiments produce positive results. In other words, the majority of your roadmap rests on unvalidated bets. The GIST framework reverses this logic. Instead of committing to fixed deliverables, each team maintains an Idea Bank, a reservoir of hypotheses scored by ICE (Impact, Confidence, Ease), then tests the best ones through step-projects of 10 weeks maximum, in the Build-Measure-Learn spirit of the Lean Startup. Think of GIST planning as a GPS that recalculates the route in real time rather than a road map printed six months in advance. Goals remain stable over one to two years and often rely on the OKR system. Ideas live and die in the Idea Bank based on collected evidence. Step-projects advance in short cycles. Tasks change every week. Each layer operates on its own time horizon, allowing decision-makers to maintain strategic direction while continuously adapting execution. Itamar Gilad details this approach in his book "Evidence-Guided: Creating High-Impact Products in the Face of Uncertainty," which has become a reference for product teams that want to steer by evidence rather than opinion.
Objectives
- Prioritize work
- Structure development
- Ensure strategic alignment
- Foster innovation
- Ensure strategic alignment
Used by
- -Google (birthplace of the framework by Itamar Gilad, notably on Gmail growth)
- -Booking.com (adoption of GIST principles for product planning, endorsed by Joseph Levin, Director of Product)
Advantages
- Drastic waste reduction. By testing ideas before committing, you avoid building features nobody uses. Google/Bing data shows fewer than 1 in 3 ideas generate a positive result.
- Permanent strategy-execution alignment. Goals give direction, Tasks follow. No gap between what leadership wants and what the team builds.
- Flexibility without chaos. Unlike a fixed roadmap, GIST planning allows course changes at the Ideas and Steps level without questioning long-term objectives.
- Democracy of ideas. The Idea Bank open to everyone prevents only managers or senior profiles from steering the product roadmap.
Limitations
- Long-term planning hard to communicate. Stakeholders used to roadmaps with precise dates can be unsettled. Prepare internal education work.
- Requires an experimentation culture. If your organization does not tolerate failure or cannot measure its results, GIST planning runs idle.
- Subjective ICE scoring. Without reliable data, Confidence and Impact scores remain estimates. Itamar Gilad's Confidence Meter helps structure this evaluation, but bias risk persists.
- Initial setup investment. Building the Idea Bank, defining Goals, training the team on ICE scoring: startup takes time before bearing fruit.
How to apply GIST Planning
- 1
Define your Goals
Formulate 2 to 5 measurable objectives aligned with the company strategy. Each goal should have a 1-to-2-year horizon, a quantified success indicator, and a clear link to the product mission. The OKR format works well here. Output: list of goals with target metrics.
- 2
Build your Idea Bank
Open a spreadsheet or dedicated tool and start collecting all product ideas without filtering. Every team member, from developer to salesperson, can submit an idea. Do not delete anything: the Idea Bank is a permanent reservoir that can hold hundreds of entries. Output: a centralized base accessible to the entire team.
- 3
Link each idea to a Goal
Review your Idea Bank and associate each idea with one or more Goals. Orphan ideas (with no link to an objective) are put on hold. This filter naturally eliminates "pet feature" projects that serve no strategic objective. Output: Idea Bank annotated with Goal-Idea links.
- 4
Score ideas with ICE
Assign each idea a score on three axes: Impact (what effect on the Goal's metric?), Confidence (what level of evidence do you have?), and Ease (what effort to test?). Multiply the three scores to get a ranking. Be honest about Confidence: a brilliant idea without proof remains a gamble. Output: Idea Bank sorted by ICE score.
- 5
Select ideas to test
Take the 2 to 4 top-ranked ideas per Goal and move them to the testing phase. Do not commit to more ideas than your team can test simultaneously. The discipline here is to resist the temptation to launch everything. Output: shortlist of validated ideas for experimentation.
- 6
Design Step-projects
For each selected idea, define a step-project of 10 weeks maximum. Each step-project is an experiment with a clear hypothesis, a success metric, and a lightweight format (prototype, MVP, A/B test, fake door). The goal is not to deliver a finished product; it is to collect evidence. Output: step-project brief with hypothesis, metric, and duration.
- 7
Execute by sprints (Tasks)
Break down each step-project into weekly tasks managed in your usual agile tool (Jira, Linear, Trello). Tasks are the only level the development team handles daily. No process change for developers: GIST planning structures strategy, not technical execution. Output: classic sprint backlog.
- 8
Measure and decide
At the end of each step-project, analyze the results. Three possible outcomes: the idea works (invest more with a more ambitious step-project), the idea is inconclusive (pivot the test), or the idea fails (return to the Idea Bank with a revised lower Confidence). Output: documented Go / Pivot / Kill decision.
- 9
Update the Idea Bank and iterate
Feed the learnings back into the Idea Bank: adjust ICE scores, add new ideas born from experiments, archive invalidated ideas. This continuous cycle ensures your planning refines over time rather than becoming static. Output: updated Idea Bank.
- 10
Review Goals quarterly
Every quarter, evaluate progress toward each Goal. Goals are rarely achieved in a single cycle. Adjust target metrics if the context has changed, but avoid pivoting Goals too often: their stability is what gives coherence to the entire system. Output: documented quarterly review.