One name in. Who should its partnerships team call first?
Every number comes from Qloo, the agent decides what to measure next and why, and every name and number in the answer is checked against the step it cites.
Every number comes from Qloo, the agent decides what to measure next and why, and every name and number in the answer is checked against the step it cites.
2 of 4
A real deal, replayed blind: given Coca-Cola and eight brands without knowing which was real, the agent called Converse first. Across 23 real deals, the real partner made the top three 17 times, and 17 times again with the wrong seed brand swapped in, so the replay measures how the candidate lists were built.
Every number comes from Qloo, the agent decides what to measure next and why, and every name and number in the answer is checked against the step it cites.
First call: Converse.
The call sheet is the one page you send your boss. Print it, or save it as a PDF.
Venues in Atlanta that Coca-Cola and Converse fans both favor: Cibo e Beve, The Blind Pig Parlour Bar and Noodle.
Fame can't tell these eight apart. Taste can.
fame (Qloo popularity) taste (affinity with Coca-Cola's fans)
| The eight | Fame and taste, 0.5 to 1 | Taste |
|---|---|---|
| Converse * | 0.85 | |
| Vans * | 0.82 | |
| DC Shoes | 0.80 | |
| Yves Saint Laurent | 0.78 | |
| Etnies Skateboarding | 0.72 | |
| HUF | 0.70 | |
| Off-white | 0.69 | |
| A Bathing Ape | 0.69 |
* the agent's calls.
The other direction (each one's fans' rank of Coca-Cola) was measured and doesn't separate these 8, so there's no table. How that's decided
1. ConverseFirst call
Converse sits closer to Coca-Cola fans than any other candidate at 0.85. step 1step 3
2. Vans
Vans is in the clear three at 0.82. step 3
Overlap check: it ran, and no two contenders stand out as reaching the same audience. step 4
Each step is one measurement the agent chose to make in Qloo. Open one for what it asked and why.
I need to confirm which brand this id is before scoring its candidates.
Fork: Qloo knew the name, so the agent went on. An ambiguous or unknown name stops here.
I need the case's fixed eight candidate brands to score under Coca-Cola's fans.
I need to score all eight candidates under Coca-Cola's fans to see whether a clear three stands out.
I need to check whether two of the contenders reach the same audience before calling them together.
The third slot is tied between DC Shoes at 0.80 and Yves Saint Laurent at 0.78, so neither is called. step 3
All eight candidates sit in one fame band, so fame does not separate them. step 3
Two-way ranks came back empty for every candidate. step 3
The id resolved to Coca-Cola. step 1
Converse at 0.85 and Vans at 0.82 lead the scores. step 3
No pair of contenders stood out as reaching the same audience. step 4
Checker rule v3.1: 14 names and numbers in the agent's answer, each matched to the step it cites.
4 steps, each measured in Qloo; every number above comes from a step. Replayed from an earlier run: 0 Qloo calls.
Every name and number above is matched to a measured step. We attacked that check with 230 altered answers; it caught 219, and the 11 that got through are listed on /audit. What got through
The counterexample: Liquid Death x e.l.f. sold out in 45 minutes with weaker measured taste. Taste is one input. See it on the audit
What this doesn't establish: purchases, visits, commercial eligibility, or whether the deal makes money.
Who this is for: the partnerships lead who has to pick up the phone, and the sponsorship seller on the other end of it.
About 6 steps, about 28 Qloo calls and 40 seconds on a name nobody has asked about yet; a name already asked replays at once.