The first 14 days of launching a supplement brand on Shopify with Claude Code

Launching MYNDR on Shopify

I'm launching a DTC brand. MYNDR is cognitive support for women in early perimenopause — specifically the brain fog that comes with it.

I've spent years running operations for other people's Shopify stores, and I write about that constantly. This is the first time in a while I'm doing it for a store I own, and I want to document it while it's messy rather than after it's tidy. So this is what the first fourteen days actually involved, what's working, and what did the heavy lifting.

Table of Contents

Before the product: where is she in her own head?

The first thing I had to figure out wasn't the product page or the offer. It was: where is this customer in her own head right now?

There's a difference between someone who knows she has a problem and is actively looking for a solution, and someone who is just starting to notice that something feels off. The second person isn't searching for products yet. She's still trying to work out whether what she's experiencing is even real.

That second person is who I'm writing to.

Pain-aware and symptom-aware are two completely different conversations. Mixing them up is why a lot of good products don't convert — the copy is answering a question the customer hasn't asked herself yet.

The ads that are working, and why

So instead of leading with the product, or even with the problem directly, the first ads just reflect her symptoms back to her:

"I can't remember words I've used my whole life."

"I'm waking up at 3am for no reason."

"I don't feel like myself and I don't know why."

No pitch. No product. Just — I see you.

Those ads are getting 10% CTR.

She doesn't need to be sold to yet. She needs to feel understood first. A quiz funnel does the rest: it meets her in that moment of recognition and walks her toward an answer, which is where the product finally shows up.

This is the part of marketing that doesn't get talked about enough. Before you sell anything, you have to know exactly where your customer is in her own journey. Meet people where they are, not where you want them to be.

What the first 14 days actually involved

Everyone talks about launching a DTC brand like it's flipping a switch. It isn't. Here's the real list for myndr.shop, grouped the way I actually think about it.

Foundation

Distribution and partnerships

Demand generation

None of this is exotic. The point is that all of it happened in two weeks, alongside a full-time business, and that only works if most of it isn't done by hand.

The stack that made the speed possible

Here's the part most people miss: none of this speed happens without AI doing the heavy lifting, and none of it needed a new tool. It's built into the stack I already had.

The store was built with Claude Code. The entire website and landing-page build for MYNDR was managed through Claude, start to finish. If you want the mechanics, the setup is the one in Connect Shopify to Claude Code — MYNDR is the store I test that on — and the Shopify AI toolkit post covers where Claude Code fits against Cursor and VS Code.

Influencer outreach runs from my inbox. Every morning, ChatGPT — connected straight to my inbox — finds ten relevant creators and writes a custom outreach message for each. I read them and hit send. No separate outreach tool, no manual prospecting.

Email, ad analysis and reporting run through Branva. That's my own company's operations layer: the email builds, the ad-account analysis and the reporting sit on top of everything above, the same way they do for client stores. Running it on my own brand is a useful forcing function — it's the loop I describe in Audit. Build. Publish. Measure. Again., pointed at myself.

Without these stitched together, the fourteen days would have been six weeks. The speed isn't hustle. It's the stack.

What I'm testing next

Symptom-first is working at the top of the funnel. Next is testing a purchase-conversion angle alongside it, to see whether the same audience responds once she's past the recognition moment — or whether that's a different audience entirely.

Meta and Google both need more runway before I trust the numbers, so I'm not going to pretend I have a ROAS to report. When I do, it'll be here, including if it's bad.

If you're launching something yourself and want to compare notes, the store is myndr.shop and I'm easy to find on LinkedIn. More about how I work is on the about page.

FAQ

Did Claude Code really build the whole store?

The website and landing-page build was managed through Claude end to end — theme work, pages, the changes that would normally go through a developer. Product formulation, brand and the offer are mine. The distinction matters: Claude did the build, not the business.

Why lead with symptoms instead of the product?

Because the customer I'm writing to isn't looking for a product yet. She's trying to work out whether what she's feeling is real. Symptom-first copy meets her there; a product pitch at that stage gets scrolled past. The 10% CTR is the evidence.

Isn't two weeks too fast to launch properly?

The list above is what shipped, not everything that will ever be needed. Email flows are basic. Ads are a first batch. The partnerships are applications, not results. The goal of the first fourteen days was to have every channel exist so each one can start collecting signal — polish comes from the measure step, not from delaying the launch.