A 50-year candle maker brought us an idea: take the art of scent distribution into the digital space. Ten months later, Scent Systems was a connected IoT product — scentcasting stations on an M-to-M cloud, an Android-and-iOS service-agent app, and an MITX Award on the wall.
For 50 years Yankee Candle had been the consumer-aisle brand for scent. The Scent Systems thesis: corporations want a brand surface beyond their logo — and the air in their lobby is the most under-used one of all.
Yankee Candle had been manufacturing candles for half a century — a consumer-aisle brand most people first met through a jar on a shelf. The thesis for Scent Systems was that the same craft could move into a different room of the business.
Corporations were starting to think of the customer experience as a brand surface — sound, light, finish, and now scent. A consistent, custom fragrance in a retail or hospitality space behaved like a logo you couldn't see, only smell.
The wedge was hardware + software together: connected diffusing devices — scentcasting stations — sitting in the spaces, talking to each other through an M-to-M cloud, controllable from a service-agent's phone. The first system of its kind in the category.
“Corporations today want to put forth a brand beyond just their logos. We can help by providing them with a specific or custom scent for their spaces.”
The hard part wasn't the hardware. It was building a UX a service agent could trust to manage a fleet of stations — every station a sensor, every sensor a setting, every setting changeable from a phone.
A service agent doesn't manage one station. They manage every station across a retail chain, every diffuser in a hospitality group, every unit in a corporate campus. The UX had to scale from "this room" to "this footprint" without rewiring how an operator thinks.
Scent Systems wasn't about flooding a space — it was about targeting. One scent at the entry. Another at check-in. A scent-free zone at retail-fixture A. The control had to be granular enough to match a brand's actual spatial intent.
The agent fleet was mixed — some Android, some iOS, some old, some new. The UI had to land identically on both: same gestures, same hierarchy, same operating muscle memory. No cross-platform behaviour drift.
The whole product is a phone-sized interface onto a fleet of connected machines — each one streaming telemetry up to a cloud, each one taking commands back down. The IA had to make that legible at a glance.
The architecture started with the service agent's job, not the product's feature list. Walk a service round. Open the app. See every station in the route, healthy or not. Drill into a station — fragrance level, liquid level, temperature, humidity, schedule. Make the call: refill, reschedule, retune the concentration, mark resolved. Move to the next station. A whole route in a coffee break.
Underneath, the M-to-M cloud was streaming telemetry from every device continuously, and the agent UI was the surface that made that flow legible. Every screen carried one live signal and one available action — no nested menus, no preference panels burying the controls. The agent never had to remember where a setting lived; the screen they were already on was where the next move belonged.
The harder design decision was what to hide. A station has a long tail of low-level controls — most of them useless to a working agent on a route. The IA's job was to surface the few that matter ninety percent of the time, push the rest into a deliberate "advanced" surface, and never let the everyday flow get cluttered by the once-a-year edge case.
“Rocket Farm did a great job working with us on the UI, to come up with one that was intuitive and straightforward to control our M-to-M cloud. Managing all the different mobile devices, from Android to iOS, and getting that consistent look and feel — it all came together really well.”
Service-agent UI is its own genre — one foot in the field, one phone in the hand, never a desktop. Three rules ran every design review.
The first thing the agent sees is whether every station is healthy. Anything off-baseline is surfaced before they have to ask. Only after the status is read does the UI offer the controls — never the other way around.
Fragrance level, liquid level, temperature, humidity — the readings that matter live on the station's main card. If the agent can't make a routing decision off the card in front of them, the card has the wrong information on it.
A fleet running on a mix of devices can't have two different operating models. Gesture vocabulary, hierarchy, terminology, status colours — all of it lands the same on both platforms. The phone in the agent's hand never decides what they can do.
The shipped product runs across native Android, native iOS, an M-to-M device cloud, and the embedded firmware on every scentcasting station. Here's the surface area our team owned.
The day-to-day surface: every station in the fleet, every reading, every available command — running natively on both Android and iOS with a unified vocabulary across platforms.
Fine-tune the intensity of any station from the phone. Set the curve over the course of a day, ramp the scent up for an event, dial it down for an empty room — all without sending anyone to the unit.
Real-time telemetry from every station — fragrance output, liquid in the reservoir, when the next refill is due. The same data that used to require a site visit is now on every agent's screen.
Temperature and humidity influence how scent moves through a space — too dry and a fragrance dies, too humid and it lingers wrong. The stations adapt their output to the room they're in, automatically.
The device-cloud backbone behind it all — every station authenticating, every reading flowing up, every command flowing down, the kind of machine-to-machine infrastructure that has to run quietly forever.
Service-agent identities, role-based access, an audit trail on every command that hits a station. A B2B operations product still has to be a B2B-grade product, even when the only consumer is the air in a room.
The Scent Systems sub-brand earned its own palette — Yankee's signature red as the anchor, warm cream as the surface, and a steel-grey supporting cast for the B2B service surfaces. A brand that reads consumer-friendly at retail and machine-room-credible at the operator level.
Scent Systems shipped on time, won the MITX Award for innovation, and gave Yankee Candle a B2B revenue surface the consumer business had never reached. The same craft, in a different room.
A working consumer-grade-feeling B2B product: scentcasting stations on the M-to-M cloud, an agent app on Android and iOS, and an operations-console surface in service in real retail, hospitality, and corporate spaces.
The first connected-scent product of its kind, recognised with an MITX Award for innovation. A 50-year-old consumer brand opened a brand-new B2B revenue line — and a product category alongside it.
The operator capability Todd describes — check every station, monitor liquid, control settings, all from a phone — was the thing the category didn't have. The fact that it's still rare is the part the team is quietly most proud of.
“A service person can check if each station is working properly, monitor liquid levels, control settings. Our competitors can't do any of that.”
If the Scent Systems story is the shape of your build — connected hardware, mobile-first operations, a fleet of devices on a cloud — here's where else our team has shown up across the portfolio.