--- title: "Get to your first evidence-backed decision" canonical_url: https://zentrik.ai/docs/product/getting-started markdown_url: https://zentrik.ai/docs/product/getting-started.md category: "Product guides" learning_track: "getting-started" last_reviewed: 2026-07-28 --- # Get to your first evidence-backed decision This path takes a new workspace from initial context to one reviewed customer learning and a clear next decision. It is for workspace owners, admins, and product teammates setting up Zentrik for the first time. ## Overview Do not try to configure every product area before the workspace contains useful evidence. Start with enough context to interpret one real source, review what Zentrik extracts, and follow that learning into the next product decision. The core model is **signal → insight → opportunity → idea → initiative**. If those terms are new, keep [From customer evidence to initiatives](https://zentrik.ai/docs/product/customer-evidence-to-initiatives) open as the product map. ## Before you begin You need: - a Zentrik workspace - an **Owner** or **Admin** for workspace settings, integrations, and invitations - one product or product area to use as the first scope - one approved source of customer evidence, such as a call, support ticket, research note, document, or pasted text - a teammate who can review whether the extracted learning is accurate Use approved test or customer data and follow your organization’s privacy rules. Do not upload credentials, secrets, or material that your workspace is not allowed to process. ## 1. Ground the workspace Open **Settings → Workspace context** and add the smallest set of facts needed to interpret the first source: 1. Describe the company and the product in plain language. 2. Add the product, relevant personas, goals, and important constraints. 3. Add or confirm the relevant product feature or product area. 4. Save names that teammates already use. Avoid creating a parallel vocabulary only for Zentrik. Context is not a place to paste every internal document. Keep durable facts and definitions here; bring source material in as evidence or a context document when it needs provenance. Use [Workspace context](https://zentrik.ai/docs/product/workspace-context) for the difference between flat context, named records, documents, goals, and product features. ## 2. Add the team Open **Settings → Team** and invite the people who will review evidence or make the first decision. - Use **Owner** sparingly for workspace ownership. - Use **Admin** for teammates who must manage settings, integrations, API keys, or invitations. - Use **Editor** for day-to-day product work without workspace administration. - Use **Viewer** for read-only access. The invitation appears under **Invitations** until it is accepted. Confirm the email and role before resending or copying an invitation link. See [Workspace administration](https://zentrik.ai/docs/product/workspace-administration) for role boundaries, billing visibility, API keys, and offboarding. ## 3. Add one real source of evidence Choose one controlled path: - connect an approved provider under **Settings → Integrations** - upload or paste a call transcript, ticket, research note, or document - use the Signals API when the source is produced by another system Start with one or a small batch. Confirm the source title, product scope, account or participant context, and processing result before scheduling a large import. [Evidence to insights](https://zentrik.ai/docs/product/evidence-to-insights) covers manual and connected imports. Engineers can use the [Signals API quickstart](https://zentrik.ai/docs/developers/signals-api-quickstart). ## 4. Review the first learning Open the processed signal in **Discovery** and check: 1. The source content is complete enough to support a claim. 2. The signal is attached to the correct product and account context. 3. Each suggested insight says something the evidence actually supports. 4. The wording separates an observed customer problem from a proposed solution. 5. Incorrect or duplicate learnings are edited, rejected, merged, or left unapproved. AI output is a proposal. A useful first result is one trustworthy learning with a visible link back to the source—not the largest possible number of generated insights. ## 5. Shape the next decision Decide what the reviewed learning needs next: - Link related insights into an **opportunity** when they describe the same problem, unmet need, or strategic bet. - Create or connect an **idea** when the team has a possible answer worth exploring. - Start a **study** when the idea needs evidence before commitment. - Promote validated work to an **initiative** when the team is ready to define scope, success, and delivery. Keep the source trail intact. The purpose of the workflow is not to move every item downstream; it is to make the next decision explainable. ## Where to go next - Build the discovery habit with [Insights to opportunities](https://zentrik.ai/docs/product/insights-to-opportunities). - Add classification only when volume requires it with [Taxonomy and themes](https://zentrik.ai/docs/product/taxonomy-and-themes). - Validate a solution with [Idea Studies](https://zentrik.ai/docs/product/idea-studies). - Commit, sequence, and communicate work with [Priorities, planning, and roadmaps](https://zentrik.ai/docs/product/planning-and-roadmaps). - Turn the decision into reviewed artifacts with [Documents and delivery](https://zentrik.ai/docs/product/documents-and-delivery). ## Expected result The first setup is useful when: - product vocabulary and goals are understandable to the teammates who use them - the right teammates have the least access they need - one approved source is processed under the correct product scope - at least one reviewed insight links back to that source - the team can state the next decision and why it is or is not moving forward - unanswered setup questions have an owner or a support handoff ## Troubleshooting ### Settings are not visible Workspace settings are available to Owners and Admins. Ask a workspace Owner to change your role or complete the administrative step. ### The first import created noisy results Check source quality, product scope, and workspace context before importing more. Review a small sample, correct the context, and only then expand the batch. ### The team cannot agree what object to create Keep source material as a signal, a supported learning as an insight, a shared problem or bet as an opportunity, a possible answer as an idea, and committed work as an initiative. ## Related guides - [From customer evidence to initiatives](https://zentrik.ai/docs/product/customer-evidence-to-initiatives) - [Workspace context](https://zentrik.ai/docs/product/workspace-context) - [Set up your product feature tree with an agent](https://zentrik.ai/docs/product/feature-tree-setup)