Your clariBI trial includes 50 AI credits. That sounds vague until you see how questions actually consume them. Here's the practical guide to spending them well.
What counts as one credit
Each question you run in Conversational Analytics, each AI-generated report, and each AI-recommended chart consumes credits. The exact count depends on the work the AI engine does behind the scenes:
- Simple lookup question (~1 credit): "What was our revenue last month?" — the engine picks one tool, runs it, summarizes the result.
- Comparison or trend (~2–3 credits): "How did MRR change quarter-over-quarter?" — the engine pulls two windows of data, computes the delta, picks the best chart.
- Multi-source or breakdown (~3–5 credits): "Which marketing channels had the lowest CAC last month?" — the engine joins ad-platform spend with conversion data and ranks the results.
- Full AI report (~5–8 credits): a multi-widget report with narrative, benchmarks, and recommended actions.
What 50 credits buys you
Realistically, 50 credits is enough for ~30–40 ad-hoc questions during a two-week trial. Most trial users finish with credits left over. The credit allowance is sized so you can explore without rationing.
If you want a rough budget:
- Days 1–5: spend ~15 credits exploring. Ask the obvious questions: what's our top product, where's revenue concentrating, which channel converts best.
- Days 6–10: spend ~15 credits on multi-source questions that no single tool answers. This is where clariBI's value lands hardest.
- Days 11–14: spend ~15 credits on the questions you'd want answered weekly. Pin the charts. By day 14 you have a dashboard of recurring answers you can keep using on a paid tier.
How to get more out of each credit
- Be specific. "Top 10 customers by revenue, last 90 days, filtered by paid plan" runs more efficiently than "show me my customers" because the engine doesn't have to ask the LLM to disambiguate as much.
- Re-run pinned charts. A chart that's already pinned to a dashboard refreshes on its own and doesn't re-consume credits — the AI work was done at pin time. Pin the answers you want to keep seeing.
- Use auto-generated dashboards as a starting point. The preprocessing-pipeline dashboards are free. They cover most of the "table-stakes" reporting and let you save credits for the bespoke questions.
- Phrase questions as questions, not commands. "How many active users do we have?" routes more efficiently than "active users". The engine spends less time inferring intent.
What happens if you run out
The Conversational Analytics interface blocks further questions and offers an upgrade prompt. Your existing dashboards, data sources, and pinned charts keep working — they don't depend on the credit balance. You can also continue with auto-generated insight dashboards (those don't consume credits).
Most trials end before the credit balance does. If you're approaching empty, that's usually a signal that you're getting real value from the AI engine — and the move from 50 trial credits to the 500-credit Starter tier is the natural next step.
The credit equation, simplified
- Trial: 50 credits for 14 days.
- Starter: 500 credits/month ($99/mo).
- Professional: 1500 credits/month ($199/mo).
- Enterprise: 5000 credits/month ($999/mo).
Customer credit pricing stays constant across underlying AI providers — so you don't have to think about which model the engine routed to. A complex question costs the same whether the planner picked a smaller or larger model behind the scenes.
See the full pricing matrix for the side-by-side comparison, or our day-one onboarding guide if you haven't started the trial yet.