Your first five steps in clariBI: a trial walkthrough

A new clariBI trial gives you 14 days, 50 AI credits, room for 3 data sources and no card on file. It is tempting to connect everything at once. A better start is one source, one question and one saved answer. These five steps do that, in order.

Before step 1: the onboarding wizard

Right after you register, clariBI opens a short onboarding wizard: pick an analysis template, connect the data it needs (or use its sample data), and see a first AI analysis. It is a good way to see the product work end to end. You can also skip ahead and connect data yourself; the steps below assume you do. The help article on creating your account and onboarding covers each wizard screen.

Step 1: connect one source, not five

Pick the tool whose numbers you already check most often. The questions you care about are about that data, so it pays off first. Common first choices:

• Stripe, as an MCP app: payments, customers, subscriptions.
• Google Analytics 4, through Google sign-in: traffic, pages, conversions. See the Google Analytics integration.
• HubSpot, as an MCP app: contacts, deals, pipeline.
• Your PostgreSQL database, with a read-only user.
• A CSV or Excel export, if your main tool is not in the catalog.

Click Data Sources in the sidebar, then Add Data Source, choose the tool and authorize it on the vendor's screen (or enter the connection details or upload the file). The first sync runs in the background. How long it takes depends on the tool and how much data it has.

The trial includes MCP apps such as Stripe and HubSpot. The Free and Lite plans do not, which matters if you plan to stay on one of those after the trial. The full list is on the MCP integrations page.

Step 2: read the dashboard clariBI built

When the first sync finishes, clariBI builds an insights dashboard for the source. Open AI Dashboards and find it. Nobody configured it: clariBI looked at the columns and chose headline numbers and charts. Building it uses no AI credits, and it stays available on every plan, including Free after the trial.

Spend a minute checking it:

• Were dates read as dates and amounts as numbers?
• Does the headline total match what you see in the source tool?
• Which breakdowns did it pick, and which is missing?

If something looks wrong here, fix it now. Every later answer depends on the same data.

Step 3: ask one question

Open Ask clariBI, click New Session, tick the source you connected and create the session. Then ask something you actually want to know and would normally need a spreadsheet for:

• "What was revenue in each of the last six months, and the change from month to month?"
• "Which 10 customers paid us the most this year?"
• "Which traffic sources brought the most sessions last month, and how did their conversion rates compare?"

clariBI plans the analysis, computes the numbers with regular analytics code on your data, and writes an answer with a chart. Every number in the text is checked against the computed results. Click a chart point to open the drill-down and see how a value was calculated.

Each question uses AI credits: at least 1 per AI call, more for questions that read more data or need a longer answer. Your balance is in the sidebar under Usage This Month. The first question guide has more on phrasing.

Step 4: keep what was useful

If the answer is worth keeping, click Save as Report. It appears under Reports, where you can download it as a PDF. Then ask a follow-up in the same session ("Break that down by product", "Which month was lowest?"). The session remembers the previous question, so follow-ups are cheaper than starting again.

If the answer is something you want to see every week, build it into a dashboard instead: AI Dashboards > Create Dashboard, pick the source and describe the metric in plain language.

Step 5: add a second source

Now connect a second tool, ideally one whose data you rarely look at next to the first. Revenue next to ad spend. Traffic next to sign-ups. Support tickets next to churn. Questions can combine sources by date or by a shared column, which is where you get answers no single tool gives you:

• "Compare weekly ad spend with new customers for the last 12 weeks."
• "Did weeks with more support tickets have more cancellations?"

That is the core loop: connect, check, ask, keep. Everything else in clariBI builds on it.

What to leave for later

• Connecting everything on day one. Each source you add to a session makes the question broader and the answer more expensive. Add sources as questions need them.
• Rebuilding every existing report. The trial is best spent on questions you cannot answer today.
• Burning credits on broad questions. "Tell me everything about my data" costs more and says less than three specific questions. Making the most of your 50 trial credits has a plan for the two weeks.

Forecasting, team workspaces, public links, scheduled reports and the REST API are not part of the trial; they start on the paid plans. When the trial ends without an upgrade, the account moves to the Free plan and keeps its data, sources and their insights dashboards. What happens when your trial ends has the details.

Start with step 1 today: connect the tool you check most, open its dashboard, and ask it the question you asked a colleague last week.