A Plan holds the objective and its context, so every step that follows (interviews, processes, findings, initiatives and the final deliverables) stays pointed at the same goal.
You build a Plan visually, as a connected flow of stages: Configuration, Discovery, Processes, Insights, Initiatives and Deliverable.
The point of a Plan is to answer the questions that actually matter to you, one at a time and in order. You scope a stage around a real question, and Horizon's AI does the work behind it: it interviews your team, reads what comes back, and turns it into insights and initiatives you can act on.
Because each card links to the real work in Horizon, answering one question feeds straight into the next, so you move forward instead of collecting another backlog. When you reach the Deliverable, the question you opened with is answered and everything that led to the answer is right there.
Every Plan opens with Configuration: your objective and the context around it. It takes a few minutes and it's worth doing well, because everything downstream reads from it.
Once it's saved, the Superconsultant carries that context everywhere. You'll see a Plan context chip in the chat as proof.
A Discovery Cycle is an AI interview campaign aimed at a group of collaborators with a defined goal. It's the main way information enters Horizon.
From the Discovery card, answer three seed questions and send them to the Horizon team. Then attach an existing cycle or hit 'Create new', which opens the Superconsultant with /create-discovery preloaded and links the cycle back to your Plan.
One catch: collaborators have to exist before you can interview them, so load them first, one by one or through Excel import. You can preview everything with Simulate before a single invite goes out.
When interviews come back, Horizon can turn them into process documentation. You provide the names and rough structure of your processes. That helps you structure each one, see how confidently it can be automated for efficiency, and compare processes so you know where to dig deeper or keep exploring.
Select the relevant conversations, and the AI writes a detailed description of each process based on what your teams actually said. Every process stays traceable to the conversations that evidence it.
What you get
Every process Horizon writes lands in the Process Library: a living catalog of how your company actually works, each entry traceable to the conversations behind it.
Explore the Process LibraryRun Insight extraction over the evidence your discovery collected. Horizon extracts structured findings, scores each one by impact and effort, and keeps them traceable to the real conversations behind them.
This is the point where raw interviews turn into something you can prioritize.
What you get
Every finding lands in the Insights Dashboard: scored by impact and effort, ranked so you know where to act first, and traceable to the conversations that surfaced it.
Explore the Insights DashboardTake the insights worth acting on and turn them into initiatives: concrete projects with owners, metrics and implementation plans. The Plan tracks them as they move, so you always know which ones are advancing and which ones need a push.
What you get
Every initiative lands in the Initiatives Dashboard: owners, metrics and implementation plans in one place, tracked as they move so you always know what's advancing.
Explore the Initiatives DashboardGenerate the deliverable that answers the question your Plan was scoped around (a presentation, a document, a report) and attach it. You walk away with a clear answer you can share, and every step that got you there is still right there if anyone asks how you arrived at it.
At any point during the Plan, you can ask the Superconsultant about anything it has produced. Paste one of these into the chat, or write your own.
What are the top 3 improvement opportunities in this plan, and why should we prioritize them?
Which findings in this plan have the highest business impact and the lowest effort to implement?
Can you summarize this plan for executives? Include the main risks, opportunities, and recommended next steps.
Every answer cites the real conversations, insights and files behind it, and you can generate deliverables in PPT, Word, Excel, PDF or BPMN straight from the chat. It answers in Spanish, English or Portuguese, whichever you use.
Read more about the SuperconsultantHorizon also exposes your data through MCP. That means you can plug the evidence behind your Plan into Claude, Cursor, or any other MCP-compatible tool and keep working from wherever you already are.
Set up MCPA Plan keeps the objective, the evidence, the decisions and the final deliverable in one place. You see the progress yourself, instead of waiting for someone to report it.
Every recommendation traces back to the conversations behind it. You can trust what you're acting on, and defend it when anyone asks.
Keep exploring
Jump back to the main platform overview, read the discovery framework behind Horizon, or see how our engineering team makes AI execution work in practice.
See how discovery, prioritization, and delivery connect across the full product experience.
View platformRead the framework behind continuous discovery, stakeholder interviews, and evidence-backed transformation plans.
Read guideLearn how Horizon made its codebase easier for AI tools to understand, plan, and implement against.
Read articlePlans are already available in your account.
Create your first Plan