The Problem
An expensive rollercoaster of stockouts and lost rank.
Dr Vanilla was impacted by stockouts and declining search rank. Inventory miscalculations led to products going out of stock, which reduced organic visibility. To recover growth, ad spend was increased, performance improved temporarily, and then inventory sold out again, repeating the cycle with increasing costs.
The underlying issue was in the backend. A transition from FNSKU to manufacturer barcodes created duplicate SKUs, which fragmented inventory tracking across the catalog. As a result, forecasting relied on incomplete data, and restocking decisions were not fully informed.
Diagnosis
Chronic stockouts on top SKUs
SKU duplication from barcode transition
Fragmented inventory visibility
Inaccurate demand forecasting
Ad spend wasted recovering rank
Organic rank evaporating with each crash
WHAT WE DID
Foundation first. Then activation.
Most agencies skip inventory work and jump straight to ads and creative. We started with the boring fix because nothing else compounds without it. Once the foundation was stable, we activated the growth motion on top.
PHASE 1 — THE FOUNDATION
1
Audited the data and rebuilt at the UPC/ASIN level
We ran a full audit of backend data and SKU structure, identified the duplication issues created by the FNSKU-to-manufacturer barcode transition, and rebuilt inventory planning at the UPC/ASIN level to unify demand and stock signals across the catalog.
2
Built statistical demand forecasting
We implemented statistical forecasting models to establish baseline demand, then layered in manual demand planner adjustments to account for promotional strategy, seasonality, and non-historical factors that models alone can't see.
3
Set product-level stocking policies
We established restock frequency, lead times, safety stock thresholds, and case pack constraints at the SKU level — and allocated deeper safety stock to top-performing products to protect critical momentum.
PHASE 2 — GROWTH ACTIVATION
4
Optimized listings, content, and SEO
With reliable in-stock rates, we layered in listing optimization. Enhanced imagery, copy, A+, and SEO refinements were made to make the listings ready to convert any traffic brough to the listings.
5
Scaled advertising on a stable base
Once inventory was stable and listings were optimized, AI-driven advertising was introduced to scale spend on SKUs that could sustain demand. This eliminated wasted ad spend on products that would go out of stock shortly after gaining traction, while helping drive sales volume back up in a more consistent and sustainable way.
VISUAL PROOF
The rollercoaster, then the climb.
The chart shows sales swinging wildly for one of many example ASINs in the months before our partnership, then stabilizing into a sustained climb. Lumian AI needed to stabilize inventory then
Inconsistent ads and inventory broke the Amazon flywheel. Lumian AI stabilized stock and ad spend, restoring healthy, consistent sales that continue to ramp.
RESULTS
Foundation in. Flywheel running.
Days Out of Stock
Inventory health restored
Foundation
Total Sales
90 days vs. 90 days prior
Growth
Sessions
Traffic followed the inventory
Discovery
Organic Rank
Keystone product — organic vanilla extract
Compounding
