# Marcus Delgado, Director of Operations at Trailhaus Gear — read of Inventory AI, May 10, 2026

> 9 years in e-commerce ops, currently wrangling a $7M DTC outdoor brand across Shopify, Amazon, and a wholesale channel, using Extensiv and a Google Sheet I've rebuilt four times.

## How I got here

Someone in the Merchant Operations Slack (the one with 4,000 members, not the other one) dropped this link with zero context, just "thoughts?" I clicked because I was already annoyed at Extensiv's support response time and was half-looking for a reason to evaluate alternatives. It wasn't a sponsored post. It wasn't a cold email. It was a peer forward, which is the only reason I gave it more than eight seconds.

## What I clicked first

The hero subhead got me: "Your inventory is either out of stock or bleeding margin through overstock." That's exactly right. Not overly wordy. Not trying to be clever. I kept reading.

Then I hit "Sync every channel instantly. Forecast demand accurately. Reorder automatically." and I slowed down. That's the same sentence every tool in this space has written for the last five years. The hero earns attention and then immediately spends it on boilerplate.

## Where I paused

At the bottom, buried below the pricing table, in small text: "no live customer revenue claimed" and "probability of meaningful success around 14%, by Fermi heuristics."

I had to re-read that twice. A product page telling me upfront that its own estimated probability of success is 14%. That's either the most disarming thing I've read on a SaaS page or a signal that this isn't actually software yet. I scrolled back up looking for "beta" or "waitlist" language. Didn't find it clearly. But then I saw "Unlock the dossier for $5" and "Adopt the build $99-$199" and "the team that built this to install, customize, and run launch with you." This isn't a shipped product. This is a business idea being sold as a concept kit. The page is dressed like a SaaS homepage but it's actually a pitch deck with a checkout button.

## What I distrusted

"Results Ops Teams See" followed by: 40% fewer stockouts, 30% less overstock, 8 hours saved per week, 15% improvement in gross margin. No source. No customer name. No time range. No sample size. Those numbers didn't come from customers because there are no customers yet. They came from whoever wrote the copy. "Set up in 10 minutes" is on the same level. My last inventory tool took six weeks to get stable syncing across two warehouses.

Also, "Machine learning models predict what will sell this week based on your sales history, seasonality, and market trends." Every single tool says this. Linnworks says this. Skubana said this. The question is always: how does it actually perform on a SKU with 14 months of history and a supplier who ships late half the time? That's never on the page.

## What would convince me

I don't need more case studies. I need one honest walkthrough of what happens when the forecast is wrong. Show me a SKU where the AI predicted 200 units and they sold 80, and explain what the tool did next. Did it flag it? Did it auto-adjust the next reorder down? Did it learn? Every tool works in the demo. I want to see the failure state.

I'd also want to see a real ops person's export from the reporting module with actual margin-impact numbers, SKU turnover rates, and what "dead stock alert" actually looks like in the UI. Screenshots of the dashboard would help. There are none on this page.

## What I'd ask in an email reply

1. The page says "no live customer revenue claimed" -- so is this software I can actually install today or am I buying into a build that's still in progress? What's the real status?
2. How does your demand forecast handle SKUs with fewer than 90 days of sales history, or products with major seasonal spikes that only happen once a year?
3. What does supplier integration actually mean in practice -- do you connect via EDI, API, or is it manual CSV upload on the supplier's end?

## Verdict: on-the-fence

The problem framing is accurate and the pricing section is more transparent than most. But the page runs a bait-and-switch: it looks like a live SaaS product and it's actually a concept kit from a studio. That's not necessarily bad, but I'd want to know that up front, not find it in the fine print after I've already read the whole page twice.

---
*Memo by skeptic persona, generated 2026-05-10. Studio breaks own self-grading loop.*