How Tree Hut turned customer feedback into product decisions with AI (and 430% more social engagement)
Tree Hut didn't bolt AI onto the process. They put it inside the conversation. By analyzing comments, DMs, and replies across TikTok and Instagram, the team built a living dataset of what customers want next - scents, formats, collabs, and timing.
The shift started with faster responses. It became a product signal engine. Since rollout, social engagements are up 430% year over year, and launch decisions are grounded in quantified demand, not wishful thinking.
The system behind the signals
Tree Hut uses an AI-powered community management tool to ingest and tag social interactions. Mentions are tied to specific scents, formats, and requests, then aggregated into a searchable database.
Over time, recurring requests turned into trackable demand. That meant product, media, and community teams could speak a common language: volume of asks, sentiment trend, and confidence score by SKU or concept.
Case study: Cinnamon Dolce goes from seasonal favorite to multi-format launch
Cinnamon Dolce lived as a vault scent, brought back around the holidays in the Shea Sugar Scrub. Fans pushed for a national, year-round return and more formats. With AI, the team could finally quantify it - thousands of specific requests tied to scent and form.
The data lined up with the spring window. Tree Hut expanded Cinnamon Dolce into Hydroglow Body Lotion and Foaming Gel Wash, and even featured it at the end of their Super Bowl spot. As Sarah Craig put it, "That moment really exists because of the findings we gleaned from our community through AI."
Faster read on launches and collabs
Instead of waiting for sell-through, the team tracks day-one sentiment and request patterns. That helps them course-correct in the next cycle before the market makes the decision for them.
Example: a fall collaboration with Peanuts surfaced a clear ask - more customized, collectible units unique to the partner. That signal now informs the level of exclusivity and packaging investment for future IP collabs.
Results that product teams care about
- Demand clarity: A ranked backlog of customer-requested scents, formats, and collabs - tied to volume and sentiment.
- Faster loops: Real-time readout on launches before sales data hits.
- Resource focus: Confidence to expand proven favorites into additional forms.
- Community impact: 430% YoY growth in engagements after implementing the AI tool.
Steal this playbook (for product development)
- Instrument the inputs: Centralize comments and DMs across platforms. Create a tagging taxonomy: scent/theme, format, size, price, collab partner, region, and urgency.
- Quantify demand: Set thresholds for action (e.g., X mentions/week for four weeks, with Y% positive sentiment). Deduplicate requests and weight by recency and account credibility.
- Make it part of the PRD: Add a "Community Evidence" section to every brief with mention counts, representative quotes, and sentiment. Greenlight gates require this proof.
- Track launches day 0-30: Monitor sentiment shifts, objections, and feature requests. Pre-agree on pivot options (pack sizes, scent variants, channel exclusives).
- Design small, testable bets: Limited drops and collab-specific collectibles to validate appetite before scaling.
- Guardrails: Human-in-the-loop review, bot filtering, privacy compliance, and clear escalation paths for sensitive issues.
- KPIs that predict sales: Qualified mention volume, positive sentiment delta, save/comment ratio, waitlist signups, and creator-led demand signals.
Tooling notes
Choose tools that can do entity linking (tie comments to specific SKUs and attributes) and push alerts into your workflow (Slack, Jira). If your team is new to this, a primer on sentiment analysis helps align methods and expectations.
What's next for Tree Hut
The "Cinnamon Dolce" approach is the template: turn community signals into focused offerings, then build experiences around them - from IRL activations to content and entertainment.
As Craig notes, the value isn't just speed. It's confidence. The team can justify expansions, placements, and creative decisions because customers asked for them - at scale, in their own words.
Level up your team's AI fluency
If you're building this capability and want structured upskilling for product roles, explore curated programs here: AI courses by job.
Your membership also unlocks:
Fans asked, AI listened: Tree Hut turns comments into products-and a 430% jump in social engagement