Purchasing Order
3000
manually handled POs/year
Bulk
PO generation in minutes
Automating PO Distribution
Before
The merchandising team exported purchase orders from Momentis to local machines, then manually attached and emailed thousands of POs to factories. Each PO took approximately 30 seconds of human handling, all compressed into a tight 1–2 week seasonal window. The process was repetitive, error-prone, and created an intense operational crunch twice a year, even though the calendar deadlines themselves could not move.
After
With Ogvi, the company can generate complete PO packages in bulk within minutes. Instead of downloading and attaching individual files, the team triggers Ogvi to create all required POs for a supplier and send them in a single batch, reducing the effective manual workload for a full PO cycle by approximately one week. The PO calendar remains unchanged, but the operational burden is dramatically lighter, freeing the team to focus on higher-value decisions such as demand forecasting and inventory strategy rather than administrative work.
Forecasting Missed Demand
model-estimated demand upside on key styles
Ogvi’s demand prediction model quantified how much revenue the brand may be leaving on the table by under-buying specific style–color–size combinations.
Before
The company relied on a mix of basic math, in-season performance, and intuition to guide buying decisions. The team knew they routinely under-bought and missed the final 25% of potential sales, but had no model-driven, SKU-level view of how much demand they were failing to capture. Decisions were constrained by what had already sold rather than a forward-looking estimate of true demand by style, color, and size.
After
Using approximately 12 years of historical sales data from Momentis, Ogvi’s experimental model now produces demand ranges (lower, median, and upper) for each style–color–size combination. In backtests on 2024 data, the model indicated that certain SKUs could have supported roughly 9–30% higher sales than what was actually sold, turning a vague sense of under-buying into concrete, per-SKU missed-demand estimates. The team plans to run the model live in the upcoming buying season, using forecasts to inform purchase orders and, if validated, translate this quantified upside into measurable incremental revenue.
Run forecasting on live orders
Move Ogvi’s experimental demand model trained on approximately 12 years of sales data and backtested on 2024 performance into live use as the company enters its next buying season. Using real order flow from early July, the team will compare predictions against actual outcomes and refine future buying decisions, particularly around the final 25% of sales currently being missed.
Build Ogvi as an operational layer over Momentis
Keep Momentis as the system of record while executing key workflows, such as PO creation, inside Ogvi and synchronizing that data back into the ERP. This allows both systems to run in parallel until Ogvi’s workflows are fully proven and ready to take on a larger share of day-to-day operations.
Revisit and upgrade the B2B ordering portal
Return to the paused Transa/Trends initiative to replace the rigid vendor-owned portal with a more tailored, data-aware B2B experience for sales representatives and boutiques. The goal is to support bulk ordering, in-season reorders, and product recommendations driven by each store’s purchasing history and current inventory.
