By
A case study on solving the forecasting blind spot in Magento (aka. Adobe Commerce) + Skubana operations.
Summary
A multi-million-dollar jewelry and piercing supply wholesaler running Magento with Skubana for inventory management spent years buying inventory against the wrong reference point. Their ERP could show them what was selling now. It could not show them what last year looked like at the same point in their seasonal cycle. The result was a constant pull between two failure modes: tying up cash in stock that would not move, or running out of the items customers expected to find.
After connecting IdealData to their store, the buyer gained year-over-year visibility inside every reorder recommendation. Within the first weekend on the refined system, IdealData surfaced reorder items their automated purchasing had missed entirely.
The reaction, in the customer's own words:
"This is what we think in our brains, but I've never seen it on paper."
This is the story of how a hands-on business owner stopped flying half-blind, and what changed when their forecasting data finally caught up with the way they actually think about their business.
The business
The company sells body jewelry and piercing supplies, primarily wholesale to piercing studios across North America. The catalog runs to roughly seven thousand active SKUs, and leadership estimates that around seventy percent of those are core products: items the industry expects every supplier to carry, where running out means watching a customer walk to a competitor.
The technology stack is typical for established Magento merchants of this size: Magento for storefront and customer data, Skubana for inventory, purchase orders, and replenishment automation. Skubana already creates draft purchase orders based on recent sales velocity. The owner's role is to review, adjust, and authorize them.
On paper, that sounds like a solved problem.
In practice, it is not.
The forecasting blind spot
The owner described the problem in a way that will be familiar to anyone running a seasonal business on Skubana, NetSuite, Brightpearl, or any of the other mid-market ERPs that anchor reorder logic to recent velocity:
"Traditional reporting only looks at the last six or eight weeks. It tells me where my velocity is right now. What it doesn't tell me is what last year looked like at this same point. That's the part that matters."
The gap is not technical. It is conceptual. Skubana, like most mid-market ERPs, treats inventory as a continuous flow problem. It watches the recent past, projects it forward, and tells the merchant when stock will run out at the current rate.
For a seasonal business, that model is structurally wrong half the year.
The company's sales run hot in late spring through summer, slow through September, then peak again in October and November around Halloween and Friday the 13th promotions. May, when they spoke with IdealData, is a slow month. A Skubana reorder based on May velocity will systematically under-buy for June. A reorder based on September velocity will systematically under-buy for October.
The buying team had been compensating for this for years, manually:
"My budget, dollar-wise, should be comparable to what I spent last year, not what I spent last month."
"I tried setting Skubana to look at the last twelve months. We ended up overspending."
So twelve-month context lived in the owner's head, not the system. Every buying decision required either trusting Skubana's recent-velocity recommendation and accepting the seasonal error, or sitting down with reports and reconciling against the prior year by hand.
Neither option scales.
"I don't have the time to sit there and go back to last year and see what we did."
This is the first half of the blind spot. Here is the second half.
The items that fall through the cracks
Skubana's automated PO generation is a good baseline. It is also, by the owner's own assessment, incomplete:
"The biggest value I'm getting from this is finding the stuff that's technically not already on the order."
Items get missed for ordinary reasons. A new product comes in but does not yet have enough sales history for the ERP to compute a meaningful velocity. A SKU with low recent velocity has a single wholesale customer who buys two hundred units at a time on an irregular cadence. A core item gets de-prioritized because its three-week trailing average looks fine, even though year-over-year, the merchandising team knows demand is about to spike for a known holiday.
The owner described a specific category of pain that any wholesaler will recognize:
"A new item comes in stock, and for whatever reason it gets missed on photography, missed on being enabled on the site. The worst part is, you've spent the money to bring it in and develop it, and then you didn't market it."
And the broader frustration:
"I've tried to put safety nets around everywhere I possibly can. Stuff still slips through."
This is the structural problem with passive inventory systems. They show the team what is in front of them. They do not show the team what has been missed.
What changed
The company connected their store to IdealData. The platform runs as a layer on top of Magento and Skubana, watching inventory movement, sales velocity, and purchase order activity in real time, and producing signals when something needs attention.
Two things were different from day one.
First, every reorder recommendation arrived with twelve-month context built in. Not "your velocity is dropping this week," but "your velocity is at one hundred and twenty-two percent of where it was at the same point last year, and last year you sold X through the same window." The buying decision no longer required mentally reconciling the ERP's recent-velocity view against seasonal intuition. The seasonal intuition was already in the recommendation.
The owner's first reaction on seeing this was the line that opens this case study:
"This is what we think in our brains, but I've never seen it on paper."
And a few minutes later:
"Given the seasonality of our business, having the reference to the year before is huge."
Second, IdealData surfaced reorder candidates that Skubana's native automation had not picked up. The owner spent a Saturday going through the first batch and described the result:
"It managed to catch a few items over the weekend that I put on order. I was like, okay, that one's a good one."
Four to five items, in one weekend, on inventory the automated PO system had not flagged.
These are not exotic SKUs. They are core inventory the business needs to have in stock. Each one represents an order the team would have placed eventually, by hand, when a customer asked for it and the warehouse came back empty. Each one represents either a missed sale or a piercing studio calling another supplier.
A third moment worth recording, because it speaks to the deeper value of the model: the buyer pulled up a SKU that had been out of stock for one hundred and ninety-one days at some point in its history. With no recent sales data to anchor against, Skubana's velocity-based reorder logic could not produce a meaningful recommendation. IdealData's signal could, because it was anchored against the prior year's seasonal pattern rather than recent velocity alone.
The general assessment, on the call, was this:
"The information is incredible. It's a robust auto-reorder."
Why the model works
The forecasting blind spot in Magento + Skubana operations is not solved by buying a better ERP. The ERP is doing what it was designed to do. It is anchored to recent velocity because for non-seasonal businesses, recent velocity is the right anchor.
The solution is to add a layer that does what the ERP structurally cannot: hold year-over-year context, watch for the items the ERP misses, and surface them as actionable signals at the moment a buying decision is being made, not in a report someone has to pull on their own time.
That is what IdealData does. Iris, the intelligence behind the IdealData platform, watches the store around the clock and surfaces what matters when it matters. The team does not log in to find problems. The signals find the team.
For a business with seven thousand active SKUs and one person responsible for most purchasing decisions, that distinction matters more than any feature comparison. It is the difference between hunting for problems and being told about them.
What's next
The customer and the IdealData team have a working roadmap of signals to layer on top of the reorder foundation. Sell-through anomaly detection, to catch the new product launches that quietly fail to take off because something in the marketing or merchandising workflow missed a step. Overstock alerts, to identify the cash flow drain working in the other direction. Eventually, a tagging system for differentiating core inventory from non-core, so the urgency of any given signal reflects the strategic weight of the SKU rather than just the dollars at risk.
Holiday and event awareness is the active piece of the conversation right now. The customer's industry has its own calendar of high-velocity days that no generic forecasting model would know about. Iris is being trained on those dates so that recommendations issued months ahead of a known peak reflect the spike that is coming, not the lull that is happening now.
The foundation is already in place, and the way this business buys has already changed.
If this sounds like your business
The forecasting blind spot described in this case study is not specific to body jewelry, or to wholesale, or to Magento. It is structural to any seasonal business running on an ERP that anchors reorder logic to recent velocity. If you run Magento, Shopify, or a comparable storefront on top of Skubana, NetSuite, Brightpearl, Cin7, or any similar mid-market inventory system, and you have ever found yourself manually reconciling your buying decisions against last year's numbers because your system cannot do it for you, this is the problem IdealData solves.
It is also the problem most merchants do not realize they have until they see the alternative. As the customer in this case study put it:
"This is what we think in our brains, but I've never seen it on paper."
If that line resonates, we should talk.