How we use Qloo

Every fact we show comes from Qloo, and you can check it.

Qloo knows what people with similar tastes enjoy across music, shows, brands, places and more. Common Ground uses it to find what a brand and a creator have in common. Here is what we ask Qloo, the rules we follow, and one example followed all the way back to Qloo.

What we ask Qloo

The example campaign made 122 Qloo lookups and kept 151 pieces of data.

Step	Qloo endpoint	What it is for	Used in the example
Looking up names	GET /search	Finds the brands, artists, shows and podcasts named in the brief. If a name could mean more than one thing, we ask which one you meant.	5
Looking up names	GET /v2/tags	Finds the interests and themes mentioned in the brief.	4
Looking up names	GET /v2/audiences	Finds the audience groups that match how the brief describes your customers.	1
Audience · Checks	GET /v2/insights (tags)	The interests, styles and moods people with these tastes share. This is what the creative directions are built from.	8
Audience · Creators	GET /v2/insights (demographics)	The age and gender lean of the people who like each name, shown next to the audience in your brief, and of each creator's audience.	6
Related interests	GET /v2/insights (entities)	What else these people enjoy: artists, podcasts, shows, films, books and brands, with which name in your brief led to each one.	14
Related interests	GET /v2/insights (places)	Places in each market's city that fit, for local shoot and content ideas.	4
Checks · Creators	GET /v2/insights (score tags)	Checks whether a direction hangs together and suits your audience, and how well creators' themes fit.	32
Creators	GET /v2/insights (score entities)	Compares every creator's inspirations, and the audiences of creators Qloo knows, against each direction in one go, so they're ranked fairly.	40
Finding real creators	GET /v2/insights (people)	Real creators and talent whose audience shares each direction's taste: people Qloo tags as YouTubers, TikTokers or bloggers, plus the category's own talent (chefs for coffee, runners for running shoes), with their public Instagram handle.	8
Your creators	GET /entities (Instagram handle)	Matches the creators you import to Qloo by their public Instagram handle, so their audience can be checked.	0

Would it work without Qloo?

We run the same brief twice, the same way each time, except one run gets Qloo's data and the other doesn't. You see the two side by side without knowing which is which, then reveal the answer. See it on the example campaign.

One example, traced back

An insight from the example campaign, followed back to the Qloo lookup it came from.

Step 1
What we asked Qloo
GET /v2/insights
filter.type
urn:entity:brand
signal.interests.tags
home cooking (urn:tag:lifestyle:qloo:home_cooking), hosting friends (urn:tag:lifestyle:qloo:hosting_friends), Coffee Rituals (urn:tag:lifestyle:qloo:coffee_rituals), Creative Dessert Making (urn:tag:lifestyle:qloo:creative_dessert_making)
feature.explainability
true
filter.exclude.entities
Blue Bottle Coffee (22B1BC87…), Stumptown Coffee Roasters (E6BC1B43…), Khruangbin (BAF6317A…), The Bear (4C305B2F…)
signal.interests.entities
Blue Bottle Coffee (22B1BC87…), Stumptown Coffee Roasters (E6BC1B43…), Khruangbin (BAF6317A…), The Bear (4C305B2F…)
API key not shown
Step 2
What Qloo sent back
- #1Maxwell House Coffeescore 0.87
- #2Verve Coffee Roastersscore 0.86
- #3La Colombe Coffee Roastersscore 0.85
- #4Tillamookscore 0.84
- #5Huckberryscore 0.84
- #6Warby Parkerscore 0.84
- #7Le Creusetscore 0.84
- #8Eaterscore 0.84
Scores only compare within this one lookup.
Step 3
The fact we show, written by code
Qloo data
Le Creuset ranked #7 of 8 brands associated with Blue Bottle Coffee + Stumptown Coffee Roasters + Khruangbin and 5 more (affinity 0.84, this query only); the input contributing most was The Bear (0.29).
Step 4
The insight, written by the AI
Insight
Kitchenware, deli and food-media references such as Le Creuset, Russ & Daughters and Eater suggest food culture may be a comfortable shared language, in line with the interview mention of The Bear.
From the direction “The Sunday counter” in the example campaign.

Our rules

Facts come from code, not the AI

Each Qloo fact is written by a fixed template straight from the data. The AI picks which facts matter but never writes their wording or numbers.

The AI's suggestions are labelled

Anything the AI suggests is marked as an insight and must point to at least one Qloo fact. It can't use numbers, make sales predictions, or say anything about individual people. If it breaks a rule, it's sent back to try again.

Fair comparisons only

Qloo's scores only mean something within a single lookup, so we never compare scores from different lookups. Creators are ranked against each other in the same lookup. If Qloo returns nothing for a name, we say “no data”, not zero.

Three checks on every direction

Does it hang together? Does it hold up if you remove one thing from the brief? Does it suit the audience you described? Each answer is shown separately.

Privacy and fairness

Only public names and public social handles are sent to Qloo. Contact details, rates and notes never are.
A creator's inspirations describe their work, not their followers. A creator's audience lean is Qloo's aggregate data for the people who like them, never about individuals.
A creator's city is only used for practical planning and local ideas, never to judge fit.
Qloo's answers are stored privately on our server and never published.

What it can't do

If the brief describes the audience only by age, Qloo can't tell directions apart for that audience, so the audience check shows “couldn't check”.
Qloo's taste data doesn't change by city, so differences between markets come from local places only.
Trend data was flat for the brands we tried, so we don't use it.
Qloo can't filter people by country, so a suggested creator may work outside your market. Check before you reach out.
Qloo knows public creators. Smaller creators you import may not be found; they're still cast on their portfolio and price.
Qloo describes groups of people. Nothing here predicts what one person will buy or how a video will perform.