Start with one real question and leave with a next step.
Start with the question in front of you, then use the guide to make the next decision easier.
Try saying
“I am trying to make this decision. What should I look at first?”
Start with context
Bring the smallest useful example.
Look for the signal
Separate what is known from what is assumed.
Take one step
Choose an action you can review.
Product vocabulary
What it does
Product vocabulary is the list of product areas your signals are sorted into, such as "Onboarding", "Billing" or "Mobile app". Once a data source is connected, Specky classifies each new signal into a hierarchy: a product area first, then a feature and a theme where they apply. It reuses the categories you add when they fit, and creates a new one when nothing matches.
Specky also labels each synced signal Actionable, Low-value or Noise. Noise is kept out of search and clustering. The Exclude noise and Actionable only filters in Insights use these labels.
Where to find it
Product vocabulary isn't in the sidebar. Press ⌘K (Ctrl+K on Windows) and choose Product vocabulary. The same view is the Taxonomy tab in Customer intelligence. The page heading reads Taxonomy Browser.
If you see No classified signals yet, connect a data source first. There is nothing to sort until signals arrive.
How to use it
- Open Product vocabulary.
- Press Add Category, type a name your team already uses, and press Add.
- Press Add child (the plus icon) on a saved category to add a subcategory.
- Under AI Taxonomy Suggestions, press the thumbs-up to apply a suggestion or the thumbs-down to dismiss it.
- Select a category in the hierarchy to see its signals, with their source and how confident the classification was.
- Under Discovered Themes, press Add opportunity to turn a pattern into an opportunity.
You can't rename or delete categories in this view. Categories that come from signals change when signals are reclassified.
Signal quality
Open Settings → Signal Quality to see the split of actionable, low-value and noise signals for each connected source. The same page has an Override Signal Classification form for reclassifying a single signal; your override wins over the AI label.
Example
In a fictional team, everyone says "Workspace setup", but signals keep landing under "Onboarding". You add Workspace setup as a category. New signals about inviting teammates and connecting a first tool can now be filed there, and the team reads feedback in its own words.
Ready to apply this?
Start with your own product and keep the first read grounded. You'll create an account before the guided result so it can be saved to your workspace.