Fragrance trend research should identify changing customer language, formats and occasions without turning the brief into a copy of current bestsellers. Separate durable behaviour from short-lived aesthetics, triangulate several evidence sources and translate findings into testable product choices.
Write the decision before collecting trends
Define whether research must guide scent direction, pack, format, price or content. A broad search for what is trending produces attractive fragments with no decision owner. Set market, customer, channel and launch horizon so evidence can be weighted.
Use multiple evidence layers
Combine retailer assortments, search language, reviews, social conversation, trade launches and direct customer interviews. Retail data shows what is offered; reviews reveal unmet expectations; social content shows vocabulary but may overrepresent novelty. Record source date and bias.
Distinguish signal from repetition
Group observations by underlying need such as portability, comfort, projection control or refill behaviour. Count independent sources rather than repeated articles referencing one launch. Check whether the signal appears across price tiers and whether it persists beyond a seasonal content cycle.
Translate insight into a prototype
Convert each chosen signal into one controlled hypothesis: scent contrast, size, application or message. Keep the brand's exclusions visible. Test prototypes against target customers without announcing the trend label, which can bias responses toward recognition instead of genuine preference.
Set an expiry date
Attach a review date and evidence threshold to every trend-led decision. Use stock components or reversible artwork when uncertainty is high. After launch, compare sell-through, sampling conversion and repeat purchase with the original hypothesis; archive trends that generated attention but not durable demand.
Project checklist
- Named commercial decision
- Market and time horizon
- Independent evidence sources
- Bias and source dates
- Brand-fit filter
- Prototype hypothesis
- Post-launch validation