Mavera DocsStart here: your first week
A day-by-day onboarding that takes you from new account to power operator.
Complete the first-week loop
Use the main product shell from first prompt through evidence and follow-up.
Use the real sidebar pattern and open Mave.
Good afternoon
Choose a surface from the sidebar to begin.
- 1Open MaveUse the real sidebar pattern and open Mave.
- 2Choose who answersOpen the audience picker and choose a named workspace audience.
- 3Write a useful requestType the decision you need help with in the composer.
- 4Attach contextAttach the sample brief so Mave has evidence to work from.
- 5Send itSubmit the request and watch the run state change.
- 6Read the resultOpen the evidence-backed answer card instead of stopping at the headline.
- 7Ask the follow-upUse the follow-up composer to press on the most important uncertainty.
Follow this and by Friday you'll be the person your team asks how it works.
The five moves
- Describe one real decision.
- Build or choose the audience.
- Add only the context the decision needs.
- Run one modeled comparison and inspect the evidence label.
- Ask what real customer or market signal would prove it wrong.
Day 1 — talk to Mave (20 minutes)
Sign in at app2.mavera.io. Don't touch the tools yet. Just chat.
Copy-paste these, one at a time, and read what happens:
What can you do for me? Keep it short.
My business is [one sentence about your company]. Remember that.
Who do you think our hardest customer to win is? Ask me questions if you need to.
Notice two things: she remembers what you tell her (check the Context page later — it'll be there, marked LEARNED), and she asks instead of assuming. That's the whole platform in miniature.
Day 2 — build your first audience (15 minutes)
Open Audience Builder in the sidebar. Describe a customer group in one sentence using the identity-behavior-trigger pattern:
Mid-career Raleigh consumers who bank with a national institution, value convenience and trust, but reconsider when a life event or service frustration hits.
Build it. Open Talk → Chat, pick your new audience in the 👥 picker, and interview them for ten minutes. Ask what they'd never tell a survey.
Day 3 — teach it your business (15 minutes)
Context page → Upload & extract tab → drop your pitch deck or one-pager → Extract context → accept the good facts, untick the wrong ones.
Then Brand kit tab → upload your logo → watch the colors extract. Then Style & voice → describe your voice → Generate examples → edit them until they sound like you.
From this moment every answer, study, and image is calibrated to your business.
Day 4 — run something with numbers (20 minutes)
Ask Mave, in chat:
Run a study with [your audience]: which of these two messages is stronger, and why — "[message A]" vs "[message B]"? 100 respondents.
Approve the plan. While it runs, read Reading results like a researcher. When it finishes, practice the follow-up muscle:
Which respondents were hardest to convince? What would move them?
Day 5 — put it to real work (30 minutes)
Pick a live question your team actually has this week. Run one of the workshops end to end. Share the result with a colleague via the conversation's Share button.
Then set up your safety net: Usage page → set a monthly workspace cap and a per-conversation cap. Invite one teammate from the Team page.
The one habit that matters
Never start over. Every follow-up question in the same conversation compounds — Mave has the full history, the results, and your context. Operators who thread one conversation per project get dramatically better output than those who open a fresh chat per question.
Choose your track
| If you are… | Read next |
|---|---|
| A researcher / strategist | Message testing workshop, Voice-of-customer mining |
| An account / client lead | Campaign pre-flight, Working with files |
| The workspace owner / admin | Set up your team, Workspaces & billing |
| "I just want prompts to steal" | The prompt library |
Optional: hand this workflow to your technical team
Developers can reproduce the same estimate-first workflow through the API quickstart. Keep API keys server-side and preserve synthetic and grounding metadata in downstream systems.