Objectives are the foundation of your growth strategy. A good objective is specific, measurable, and focused on a single outcome.
Creating an objective
When you create an objective, describe what you want to improve in plain language. AI analyzes your description and matches relevant playbooks.
Good objective examples:
- •"Improve day-1 retention from 30% to 45%"
- •"Get more users to complete onboarding"
- •"Increase referral rate to 15%"
- •"Reduce time to first value"
Note: Avoid vague objectives like "grow the business" or "get more users" - these are too broad to generate useful recommendations.
Business context
Each project can store business context — your industry, stage, audience, and a company summary. This is set once and feeds into all AI features automatically, so you don't need to re-enter it for each objective.
Where to set it:
- •During onboarding — the optional context section saves to your project
- •In settings → Projects → click the edit button on any project
- •You can also paste your website URL to auto-detect your context
Tip: Setting your business context makes every AI feature smarter — from playbook matching to experiment suggestions to result analysis. It takes 30 seconds and is worth it.
How AI matching works
When you create or update an objective, AI analyzes your description and matches it against our playbook library. It considers:
- •The funnel stage (acquisition, activation, retention, monetization, referral)
- •The specific problem you're solving
- •Your business context (industry, stage, audience) from your project settings
- •Previous experiments you've run
Finding experiment ideas with signals
Not sure what to test? Every objective page includes a "Where to find experiment ideas" section with concrete prompts that tell you exactly where to look — your analytics, customer feedback, competitor activity, and support tickets.
How signals work:
- •Expand a category (acquisition, activation, retention, etc.) to see prompts
- •Each prompt tells you something specific to check, like "Look at your landing page bounce rate" or "Ask a recent signup: how did you find us?"
- •Click a prompt to pre-fill the signal input, then type what you found
- •Tag your signal with a source (analytics, user feedback, competitor, support)
- •Once you've logged signals, click "Get experiment ideas from signals" — AI will suggest experiments that directly address what you observed
Tip: Log signals as you notice them, even small ones. A pattern of small signals often reveals a big opportunity. "3 users mentioned confusing onboarding" is more actionable than a vague feeling.
Managing objectives
You can have multiple objectives, but we recommend focusing on 1-2 at a time. This keeps your experiments focused and results clear.
- •Archive objectives when they're complete or no longer relevant
- •Update objective descriptions as you learn more
- •Link experiments to objectives to track progress
Tip: Review your objectives monthly. As your product evolves, your growth priorities should too.