Skip to main content

Tribes FAQs

1. Which audiences (Tribes) can I build?

Today a Tribe is built from generation: Gen Z, Millennial, Gen X, and Boomer. Additional dimensions such as gender and income are marked "coming soon" and will be enabled as the underlying datasets land.

2. Which brands and categories can I analyze, and which insights work for each?

You pick a focus that is either a brand (for example, Walmart) or a category (one of the nine GICS sectors). Brand Affinity is available for every focus, brand or category. Overlap is currently available for retailer focuses only. Coverage is deepest for major national retailers and the GICS-sector categories, which return the fullest brand lists; very narrow or niche focuses can return thin results.

3. How fresh is the data?

Affinity and Overlap reflect the most recent complete month. Card data lags roughly one to two months, so the latest period fills in over time. The window is set to avoid showing a half-empty current month, so a light recent month is expected rather than an error.

4. How does billing work? Which tokens does Tribes use?

Tribes spans two surfaces, and each draws on its own balance:

  • Analyzing a Tribe in Lenses (the chat persona) returns a quick, high-level read and consumes MCP tokens, the same balance any Lenses chat draws on.
  • Exploring further in Views is where you get the complete data, and it consumes platform tokens from your wallet (balance shown top-right), a separate balance from your MCP tokens. In Views you only spend when you buy the underlying data for a configuration you haven't purchased before; data you've already bought loads again at no cost (it shows "Loaded" rather than prompting a repurchase).

The two balances are independent: a purchase in Views never draws down your MCP tokens, and a Lenses chat never draws down your wallet. You always see the price before you buy in Views, and you can't spend more than your wallet balance.

5. Where can I find the raw data behind a result?

Data you've purchased is saved to your account, so you can revisit it or collect the underlying raw data anytime from Order History.

6. What does the affinity score (NPMI) mean, and what is "over-index"?

Affinity is measured with NPMI (Normalized Pointwise Mutual Information), a co-occurrence score normalized against an independence baseline: 0 means no different from chance, a positive score means the audience is more likely than the average customer to also shop there, and a negative score means less likely. "Over-index" (Δ) is the audience's affinity minus the affinity of all the focus brand's customers for the same brand, which separates a genuine audience skew from a brand that is simply popular with everyone. See The analyses in the Overview for a worked example.

7. What insights are coming next?

Brand Affinity and Overlap are available today. Retention and Engagement are on the roadmap and show as "coming soon" until their data lands. See Other insights on the roadmap for what each one answers.