A live demand record for every product in your catalogue — views, momentum, scarcity, lifecycle. Recomputed in real time from your shoppers’ behaviour, and visible through the portal.
Book a demo→Each product in your catalogue maintains a continuously-updated demand state. View, add-to-bag and purchase counters across nine time windows. Lifecycle, momentum, rank, stock. The selection engine reads from these records. The portal reflects them. Signal queries them.
Live, not last night.
Five states classify where each product sits in its demand arc. Transitions are automatic — computed from views, add-to-bags and purchases against the product’s first-seen date. So “New in” stays reserved for products genuinely showing the launch dynamic.
Recently published. Minimal demand history.
Early traction is beginning to appear.
Strong early demand inside the new-product window.
Was trending. Momentum has dropped.
Stable demand pattern with sufficient history.
Momentum measures acceleration relative to the product’s own recent baseline — not against the catalogue average. A product going from 5 to 10 views per hour has momentum. A product steady at 100 views per hour has attention, but not momentum. The engine surfaces both as different signal families.
Attention is a level. Momentum is a direction.
The same model surfaces two complementary commercial questions. Low stock with accelerating demand — the products worth restocking. Healthy stock with weak demand — the products worth promoting. Both are ranked from the same demand state.
Per-type debounce windows keep the feed signal-only.
The same demand intelligence flows through the platform. Read by the message engine on every request. Watched in the portal by your team. Queried by Flockr Signal in natural language.
The message engine reads the current state of every product on the page and selects the strongest signal per slot. Sub-100ms, computed live.
The same records power the portal Demand page. Five tabs, updated every 30 minutes — so your merchandisers see what the engine sees.
Flockr Signal queries the same records in natural language. Ask what's losing momentum, what to restock, what to feature — answers grounded in current state.
30-minute walkthrough. We’ll show you the live demand model running on a catalogue like yours, plus a walk through the portal and the architecture.
