# Marcus Tran, Founder/Operator at Coastline Supply Co. — read of Merchant Q&A, June 22 2026

> 8 years running a DTC supplements brand out of San Diego, currently at about $900K ARR, two warehouse guys and one part-time VA handling support. I pick up my daughter from school at 3pm every day and usually do my tool-shopping during that commute.

## How I got here

Googled "AI for ecommerce customer questions" last Tuesday after a bad week where we got buried in "where is my order" and "does this contain X allergen" tickets. Clicked three paid ads, closed two of them immediately. This one was the third. It came up organic, I think, or maybe it was in a sidebar roundup someone posted in the Shopify subreddit. I bookmarked it, came back today.

## What I clicked first

The hero line "Answer Every Customer Question Instantly" did its job. That is exactly what I need. I kept reading. "No Dashboard Digging -- Your assistant learns your product catalog and answers questions directly." Okay, I've heard this pitch before from Tidio and Gorgias AI. But I'm still reading.

Then I hit the adoptability score box and stopped cold.

## Where I paused

"Honest disclosure: we don't have live customers on this idea yet. We shipped the strategy package; you ship the customer conversations."

I read that three times. So this is not a product I can install. This is a packaged business idea being sold to someone who wants to BUILD this thing. I came here thinking I was a buyer of a support tool. Turns out I'm being pitched as a potential founder of a support tool company. That is a significant mismatch and the page does not flag it until you scroll past the feature bullets and the free trial button.

The "Start Free Trial" button in the hero is for a product that does not exist yet. That's a problem.

## What I distrusted

"Reduces Support Load -- Handle 10x more customer questions without hiring." No number attached. No merchant name. No ticket volume before and after. Every single AI support tool I have looked at in the last two years says "10x" something. It has become the new "enterprise-grade."

Also: Year-1 take-home is listed as NEGATIVE $5,000. And 1 in 6 meaningful success odds. I actually respect that they disclosed this. But if I am a merchant reading this page thinking I am buying a tool, those numbers make no sense. If I am an operator thinking about building this tool, those numbers are pretty discouraging before I have even clicked anything.

The financial upside axis scored 2 out of 10 and the page still has a $99 "Adopt the build" button. I am not sure what I am supposed to do with that information together.

## What would convince me

If I am the intended audience (a builder/operator, not a merchant), I want to see one conversation thread. Not a mockup. An actual Loom of the thing routing a real ambiguous question -- "can I stack this with creatine?" -- and either answering it correctly from a product page or handing off cleanly to a human. Thirty seconds of that would do more than all the bullet points combined.

If the audience is merchants, I want one merchant name, their monthly ticket volume, and what percentage the AI handled. Even a small beta shop. Even a friend's store.

## What I'd ask in an email reply

1. The "Start Free Trial" button -- is there an actual live demo environment I can test against my product catalog, or does the trial just give me access to the dossier?

2. The scoring shows buyer clarity 10/10 but landing page quality 6/10. Your own rubric is flagging the page. What specifically did you grade down and why haven't you fixed it?

3. Who is this for? The hero reads like a merchant tool pitch but the pricing tiers read like a founder starter kit. I want to understand who walks away with what.

## Verdict: on-the-fence

The honesty about negative year-one returns and no live customers is genuinely unusual and I noticed it. But the page is pitching two different people at the same time and landing cleanly with neither. If there is a working demo I can actually touch, I would reply.

---
*Memo by skeptic persona, generated 2026-06-22. Studio breaks own self-grading loop.*