How to detect margin erosion before it kills a product

TL;DR: Revenue dashboards can't show margin erosion, because the sale still happens — only the profit on it shrinks. Merchants in PainHunt's Inventory Management data describe products that used to win now barely breaking even, with no tool that warns them before it's critical. The same category holds a lower-tech version of the same blindness: stock tracked in notebooks, discovered empty only when a customer asks. The wedge is cost-aware inventory — tracking what each unit cost, not just how many exist.

The evidence

PainHunt's Inventory Management category holds 77 high-scoring signals (10+/15), average intensity 7.6/10, average score 11.8. The source mix is distinctive: Discourse (55) dominates — merchant and seller community forums — followed by Medium (5), Reddit (5), BlueSky (4) and Mastodon (3). These are operators talking to other operators, not app reviews.

24 of the 77 signals (31%) form a cost-and-visibility cluster, intensity 7.8/10. It splits cleanly into two versions of one blind spot.

The margin version, from Shopify store owners selling physical products:

• Supplier costs increasing, causing margin erosion on winning products.
• Products that were profitable are now barely breaking even.
• No automated tools to detect margin erosion before it becomes critical.

The visibility version, from small shop owners — one signal comes from shops in Nairobi:

• Stockouts happen silently — nobody notices an item running low until a customer asks and it isn't there. The lost sale is never recorded anywhere.
• Overstocking wastes money — without clear stock visibility, owners reorder what they already have too much of.
• Human error compounds over time — wrong numbers in notebooks, miscounted restocks, forgotten sales make the data progressively less accurate.
• Two hours every Sunday spent counting stock manually instead of running the business.
• Tracking happens in worn notebooks, WhatsApp messages, or phone spreadsheets.

Both groups are missing the same thing: inventory that knows what it cost. The requested fixes say so — automated inventory tracking with cost monitoring, margin erosion alerts, supplier cost tracking over time, low-stock alerts, and a mobile-first interface usable from any phone.

Why now

Costs stopped being stable. Sourcing, freight and currency have all moved enough in recent years that "the cost we set the price from" and "the cost we're paying" drift apart within a season rather than over years. Pricing set once at launch used to survive; now it silently expires.

The tooling didn't follow. Ecommerce platforms optimize the sale — conversion, checkout, fulfillment — and treat cost of goods as an accounting concern that shows up later in a reconciliation nobody enjoys. So the merchant's daily dashboard reports revenue, which keeps looking fine, while the number that matters isn't on it.

At the small end, the gap is more basic. Phones are universal and cheap connectivity is common, but inventory software still assumes a desk, a barcode scanner, and setup time that a one-person shop won't spend. That's why the honest description of the current system is a notebook — not because nothing exists, but because nothing existing fits.

The wedge

Make cost a first-class field, then alert on the derivative.

• Cost per receipt, not per product. Store what each batch actually cost when it arrived. Everything else — true margin per sale, erosion over time, which supplier moved — falls out of that one modeling decision, and it's the one most inventory tools skip.
• Alert on the trend, not the threshold. "This product's margin has fallen 40% over two months and is now under your floor" is actionable while there's still room to renegotiate or reprice. A monthly P&L reports the same fact after the quarter is lost.
• Rank by margin contribution, not revenue. The top-revenue product and the top-profit product are frequently not the same, and the data says merchants find this out late. Making the difference the default view is a small change with an immediate answer attached.
• For the notebook segment: phone-first, thirty-second setup. No barcode hardware, no SKU taxonomy, no import wizard. Log a sale or a restock in a few taps; low-stock alerts come free once the counts are real. The competitor is paper, so the bar is speed and reliability, not features.
• Supplier cost history as its own artifact. Which supplier raised prices, when, and by how much — the record you need to negotiate. Nobody keeps it, and it accumulates value the longer the tool is used, which is a better retention mechanic than another dashboard.

Risks and honest caveats

• The Shopify app store already has inventory apps. Some do track cost. The specific gap in this data is proactive erosion alerting rather than after-the-fact reporting, which is narrower than "build an inventory app" — and narrow enough that an incumbent could add it.
• The two segments are two companies. Shopify merchants with supplier data and Nairobi shop owners with a notebook need different products, price points and go-to-market. The signals sit in one category; the businesses don't. Trying to serve both is the most likely way to serve neither.
• Cost data entry is the real adoption barrier. Margin alerting requires accurate landed cost per batch, and that's manual work at exactly the moment a busy merchant is receiving stock. If the capture step isn't nearly free, the data goes stale and the alerts become wrong — which is worse than absent.
• Small-shop willingness to pay is unproven here. A shop owner who spends two hours a week counting stock has a real cost, but converting that into a monthly subscription is a separate claim this data doesn't support. The same segment appears in the data explicitly avoiding subscriptions elsewhere.
• Discourse-heavy sourcing skews operator-sophisticated. 55 of 77 signals come from forums, so these are merchants engaged enough to post in a community. The quieter majority may have the same problem and less appetite for a new tool.

How to validate this further

Read the Inventory Management signals in the Pain Point Browser and separate the two segments before picking one — they read as a single category and behave as two markets. For adjacent commerce wedges from the same data, see ecommerce overselling and order integrity and Shopify app sunset and fulfillment continuity. Then test the cost-capture problem — the thing most likely to sink this — with the Idea Validator and how to validate a startup idea.