How Organic Intelligence gave L'Oréal a read on 75 ideas in 48 hours
Shoppers had stopped treating online and offline beauty as separate things. L’Oréal, who call it O+O, needed to define what the next generation of beauty retail should look like, and needed it fast enough to land at their annual global meeting six weeks later.
They arrived with 75 ideas and a quiet worry underneath them. Were these the right 75?
The challenge: One question, several complications
Beauty shoppers’ expectations had shifted more in three years than in the previous three decades. Personalization at scale had stopped being a differentiator and become the price of entry.
Three things made the brief harder than it looked:
- L’Oréal is not a retailer. Whatever came out of the work had to be persuasive enough to move the retailers who actually own the store.
- The answer had to hold globally while respecting local reality. WeChat conversations with in-store experts are unremarkable in China and genuinely novel almost everywhere else.
- Six weeks. Traditional concept development would have eaten most of that before anyone tested anything.
Then the quieter problem. L’Oréal had narrowed their long list to 20 hypotheses they believed in. Nobody could say whether those were the strongest 20 or simply the 20 that had survived internal debate.
Services used
The solution: Ask better questions before asking anyone anything
We started in a room. A day-long acceleration workshop in Paris with L’Oréal’s insight team and stakeholders, because the brief raised more questions than it answered.
Who is this store for?
What does good look like?
How do local teams feel about being tested globally?
And the direct one: are you confident these are the right initiatives to put in front of consumers?
That set the approach. Three stages, in sequence.
- Expand the opportunity space.
Organic Intelligence built two custom data lakes, one on beauty conversations and one on retail, drawing millions of data points from TikTok, Instagram, X, Reddit, Facebook, news, blogs, forums, and e-commerce reviews. All collected GDPR-compliantly, all built for L’Oréal rather than pulled off a shelf.
Analysts codified the 75 existing ideas and set the vision. AI then went looking past them, running 373 queries across the lakes, with analysts reviewing findings alongside it. The process surfaced 86 additional ideas, which analysts filtered for strategic fit and brand relevance.
It also surfaced two areas L’Oréal’s framework had not included at all: sustainable shopping and live commerce.
- Prioritize with consumers.
The shortlist went into
quantitative testing to size demand and rank initiatives on their potential to drive footfall and spend. - Pressure-test with retailers.
Because L’Oréal needed retailers to act, retailers had a say in what made the final recommendation.
The Organic Intelligence stage took 48 hours.
The impact: Three of the top five came from AI
75 ideas became 11 priority concepts. Four genuinely new initiatives, surfaced by Organic Intelligence, took the shortlist to 15.
Then the finding that settled the internal debate. Of the five winning initiatives in quant testing, three came from AI-led discovery. Ideas nobody in the business had proposed outperformed ideas the business had spent months developing.
Why they won:
- The inputs were broader than any internal process can reach. Millions of real conversations rather than a workshop’s worth of thinking.
- AI made connections humans would not have made from the same material.
- Analysts supplied the judgment about what actually fit L’Oréal, which is the part AI cannot do.
What the work delivered:
- Retailer-ready O+O recommendations, shared globally and built into seller stories
- A shared view of the priority initiatives across regions, with room for local teams to weight what matters in their market
- Every initiative optimized against both shopper and retailer perspectives
- Faster rollout, with local teams able to act sooner
Delivery mattered as much as the finding. A KPI alignment workshop up front, collaborative storyboarding on the narrative, then reports built for each stakeholder team rather than one deck sent to everybody.
Why it matters
Concept development usually starts with what a business already thinks. You brainstorm, you shortlist, you test the shortlist. The ceiling on the answer gets set before anyone speaks to a consumer.
Organic Intelligence raises that ceiling. It reads what people are already saying, at a scale no workshop can match, then hands the judgment back to analysts who know the brand.
L’Oréal went into testing confident they were testing the right things. Three of the winners were ideas they would never have tested.
