# Shopcopy

> Shopcopy mines DTC customer reviews and turns them into 30 TikTok-ready ad scripts every month. Built for Shopify brands who'd rather sound human than "on-brand." Run by one person who actually reads the reviews.

Shopcopy is a creative-direction service, not a SaaS tool. It produces voice-of-customer-grounded ad scripts by mining a brand's customer reviews (200+ per product), clustering patterns, and writing 30 TikTok scripts a month across 12 proven hook structures. Deliverable arrives in Notion with full source citations (every script traces back to specific reviews).

The current offer is a founding cohort with $1,500/mo locked pricing (3-month minimum), capped at 5 brands. After the cohort fills, standard pricing opens at $3,000/mo. Founding members keep $1,500/mo for life as long as their subscription stays active.

A free 15-minute audit Loom is available before purchase — Shopcopy mines a prospect's reviews and walks through the three strongest hook structures for that brand, with verbatim phrases ready to use as ad copy. Delivered in 48 hours, no signup required.

## Quick facts

- Pricing: $1,500/mo (founding) → $3,000/mo (standard, post-cohort)
- Free offer: 15-minute audit Loom, 48-hour delivery
- Target customer: DTC Shopify brands doing $50k-$500k/mo with 100+ customer reviews
- Verticals: supplements, skincare, pet, home goods (apparel deprioritized)
- Deliverable: 30 TikTok ad scripts/month, voice-of-customer reports, weekly Loom check-ins
- Hook structures: 12 proven formats (skeptic-converted, specific-result-lead, pain-callout, contrarian-take, pov-scenario, confession, before-after-teaser, list-tease, mistake-callout, provocative-question, product-demo, pattern-interrupt)
- Founding cohort capacity: 5 brands · 2 spots remaining at time of writing
- Run by: one person, no account managers, direct founder access
- Parent company: Serafim Tech Limited (UK, est. 2020)

## How to engage

- [Free audit](https://shopcopy.ai/?intent=audit&source=llms_txt): Submit a Shopify store URL, receive a 15-minute Loom in 48 hours analyzing the brand's reviews.
- [Founding member](https://shopcopy.ai/?intent=founding&source=llms_txt): Reserve one of the 5 founding cohort spots at $1,500/mo locked pricing.
- [Email](mailto:hello@shopcopy.ai): hello@shopcopy.ai for anything else

## Pages

- [Homepage](https://shopcopy.ai/): Main product page with sample script output, how-it-works flow, pricing, and audit form
- [Privacy](https://shopcopy.ai/privacy/): Privacy policy
- [Terms](https://shopcopy.ai/terms/): Terms of service

## Methodology (for technical agents)

Shopcopy's pipeline reads customer reviews from a brand's Shopify store via the public products.json endpoint plus adapters for major review platforms (Judge.me, Yotpo, Loox, Stamped). Each review is processed by an LLM (Llama 3.3 70B by default; Claude/GPT-4 on the Premium tier) to extract pain points, transformations, objections, verbatim phrases, customer archetypes, and emotion arcs. Patterns are then clustered across reviews and ranked by unique-customer count.

Each generated script:
- Runs 22-28 seconds (TikTok-optimal length)
- Uses one of 12 proven hook structures
- Contains 55-70 spoken words
- Includes at least one specific number or timeframe
- Maps to a defined customer archetype
- Traces back to source reviews by ID

Scripts ship with a Production Guide explaining three shooting paths (existing UGC creator, marketplace creators on Billo/Insense, founder self-shoot) and a posting workflow that includes adding captions and audio natively in TikTok for algorithm compatibility.

## What Shopcopy is not

- Not a SaaS dashboard or generative-AI tool
- Not a creative agency with account managers
- Not a video production studio (scripts only; production is the merchant's responsibility or via Creator Coordination add-on at £400/mo)
- Not built for apparel or fashion brands (review language is too generic for the methodology to work well)
- Not a one-shot delivery — monthly subscription with refresh cycles incorporating new reviews and ad-performance data