Brand I've built: An ecosystem of modular phone accessories that people SNAP&SWAP to match their outfit and mood.
Part OneThe photo record.
Part TwoThe decision record.
Click any image to enlarge
Part One
Photo record
Roughly chronological. Screenshots are from the working files: Notion, Figma, Google Sheets, Shopify.
Market researchJan to Jun 2024
Bought and used most of what was already selling in China and the US. This is what decided the print process, the material direction, and which gap was worth building into.
Bought competitor cases to study print processes: 3D print, cotton with acrylic, silicone, glitter, mirror. Used to pick which process we would build on.Researching patterns and design elements already on the market.Second round of factory material samples, tested against more market samples.Comparing our sample against a top-selling US brand on quality and print.Retail research in Korea.Exhibition. Built an MVE (minimum viable ecosystem) out of products bought on the market, to test Snap and Swap before we had our own.
Product design and samplingApr to Nov 2024
500 sketches, narrowed through customer discovery, then 10+ sampling rounds between the US and China across cases, frames, cardholders and the inlaid metal logo.
Product and technique design.Divergent design tests before sampling. Used to collect customer feedback and see which patterns pulled our target customer.Interior customization design. Camera bezel, chain hole, recessed logo, 2mm lip over the screen.First factory test. Sent 1 drawing to several factories and compared what came back.Trade show prep. The accessory set: magnetic, fashion, fun, functional.
Sampling the metal logo that gets inlaid into the product.Nth design iteration. Narrowing down the first launch selection.Full line sampled, factory locked. Choosing which designs launch and which removable frame colours pair with them.Testing the cardholder for magnet strength and how much it holds.Sample back with the interior print.Cardholder tooling at the factory. Our own cut and two-tone panel.
Product naming and narrative.Our product design. 3D double-layer print, and how it pairs with each accessory.Product design. The wallpapers and stickers that ship with the case.Nth round of test samples.Measuring where the logo sits on the cardholder.
The first launch lineup.
Chain and clasp developmentMay to Oct 2024
The chain ran on its own track, because the proprietary clasp system was the innovation. Product design, mechanism testing, supplier sourcing, and bead testing across materials, sizes and finishes. All of it assembled by hand.
Sourcing raw materials for the chain. High quality alloy hardware from Japan, for our own clasp design.Same trip, hardware supplier.Parts supplier, Tokyo.Testing chain construction and trying designs.Designing and trying chain styles.Testing beads for the chain on size, finish, material, price and weight.Buying beads and testing materials for the chain.Making the chains.
Sourcing and manufacturingChina and Japan, May to Jul 2024
Explored processes and materials market by market, then visited the manufacturers we work with and signed there. Alongside that, deep visits to other suppliers and material markets to learn the techniques behind each product and how the market actually prices them.
Visiting and negotiating with phone case factories.Injection moulding floor.Sourcing bead suppliers, Yiwu.Visiting the production line.Factory visit. Learning the process.Sourcing accessory suppliers.
Choosing the leather for the cardholder.Sourcing suppliers.Phone case factory visit.Buying custom bead materials.Bulk order leaving the forwarding warehouse.First bulk order lands in the US, ahead of presale.
PhotoshootBoston, Oct to Nov 2024
2 shoots, 6 products, shot in an Airbnb. Completely DIY, no agency, under $500. Each product had its own shot list, moodboard and palette before the crew arrived.
On set.On set.Renting the lighting.Testing the light myself.Directing the model.Bubble setup on set.Reviewing selects.The crew.
Product photoshoot.Second model photoshoot, for the new drop.The studio we built.Setting up the lights.Setting the light myself.Lifestyle stills, shot between setups.Product photoshoot.Reviewing selects across both shoots.
Packaging designJan to Feb 2025
A magnetic, stackable, playable packaging system, designed from scratch so the boxes lock together the way the products do. The drawer box was specified like a product: purpose, material, dimensions, magnet count and placement.
Packaging concepting.Packaging concepts.Factory spec confirmation. Magnet count and placement.
Showing the factory how the magnets should line up.Packaging concept samples.Product information card designs.First version of the product information card.Packaging production run.
Packaging concept draft.Box dieline.Drawstring pouch spec.Packaging combinations, mapped against order types and shipping.
Brand system and visual design2025
The brand definition and the naming system, the Shopify store designed end to end, product naming and narrative, Instagram post and story design files, launch story sequences, a content format toolkit, competitor content analysis, and the copywriting.
Designed the entire Shopify store.Web design, product catalog.Instagram post design.Instagram story design.
Instagram content format toolkit.Copywriting for Instagram posts and stories.Content analysis. Breaking down how other brands sequence a launch.
Brand, identity and narrative.Applying the original logo across branding concepts.
THE SKY SwapCharm launch story.Strawberry Matcha SwapCharm launch story.
Mini World SnapCase launch story.THE SKY SwapCharm, editorial and product.Product introduction stories.Product introduction stories.
A 10 day invite only presale, with no ad spend and no discounts. 56% converted, $84.63 AOV, and 85%+ of buyers posted about it themselves. Every post from pre-heat to launch was scheduled to the hour, and had a stated logic before it had an image.
Presale launch prep. Every post scheduled to the hour.Presale launch prep. The logic and the reference behind each post.
126 comments and 300+ engagements on our first post.85%+ of presale buyers posted about us on their own Instagram.
The 113 Kitchen menu, printed for presale.Every package ships with the 113 Kitchen menu.
Pre-launch marketing material.Post-presale thank you letter.Packing station.
Presale invitation, written per person.
In-house video content2025
A channel started from zero. 1 million views in 3 weeks, 10,000 subscribers in 1.5 months. Hired a VA to do the editing, $2 per video.
Filming.Editing.Editing.Editing.Video cover.Video frame with the on-screen hook.
The channel.The channel.YouTube analytics. 1 short at 6,028,000 views.YouTube analytics. 1 short at 5,051,226 views, +531k in 48 hours.YouTube analytics.13,894 subscribers.
50,000 subscribers.50.4K subscribers.
Influencer and UGC program2025
Sourcing, outreach, screening, onboarding, weekly operations and performance-based payment. Run end to end, and every step left behind a document, so the pipeline outlives whoever is running it.
Influencer tracking sheet. Performance per post.Influencer tracking sheet. Creator research.
The UGC program, in Notion.The influencer program, in Notion.
Creator collab post.
Creator unboxing.Creator showing the frames come off.Creator opening the bottom drawer.
Creator video on YouTube.What goes in an influencer set. 9 items.PR packages going out.Creator content on TikTok.Creator content on TikTok.Creator content.Reposting a customer unboxing, frame by frame.
Customer posts2024 to 2025, Boston to Seoul, Shanghai, Hong Kong, Paris
Posts and stories from customers, across 5 countries.
Part Two
Some problems, what I did, and what they left behind
01
A lobster clasp can decide your return rate
Context
The chain has a clasp of our own design. Everything the product promises runs through that one part.
Problem
Everything I could source in China was the cheap version. The clasps broke, and they would not take the weight of a phone. A chain that drops a phone is not a defect you can discount your way out of.
Solution
I went to Japan and sourced the hardware there. Better alloy, and it holds.
Insight
That trip is where the sourcing rule came from: buy by component, not by supplier. Hardware from Japan, beads from Yiwu, cases and moulds in China. Crossing a border is worth it for the part that decides whether the product survives a customer.
02
A rebrand turned a full production run into scrap
Context
The first logo was custom, cast in metal by a jewellery supplier with a high minimum. I ordered 600. By the time they arrived they were already embedded in the cases and the chains, and the packaging was printed.
Problem
Then we rebranded. The rebrand was right for everything except that inventory. I was about to launch with a logo I was retiring, that nobody had ever seen, while marketing a new one. Tens of thousands of RMB, and a story that made no sense.
Solution
So I sold it as the point. A logo nobody would ever see again is, technically, the rarest thing the brand would ever make. There were 500 of them, already sitting in packaging printed with it.
They shipped as an exclusive commemorative edition, limited to the 500 that already existed, each with a curated card telling the origin story of a logo the brand had already retired. Only those pieces get the card and that box. The presale invitation said it: you are among the first people to hold the version of this brand that will not exist again.
Insight
What is true inside the company is a constraint. What the customer experiences is a decision. Those are 2 different jobs, and the second one is where the work is.
03
I invited influencers at every follower tier. None of them converted.
Context
I ran the influencer program and invited creators at every tier. Nano, micro, macro.
Problem
Reach went up and sales did not move. Sorting by size was the mistake. Fashion creators had beautiful feeds, but a phone case is maybe 2% of a fashion frame, and their audience is there for a look, not for a product. That is top of funnel wearing the costume of bottom of funnel.
Solution
So I built a different axis. Content has 3 possible jobs: aesthetic, entertainment, or education. Any of the 3 can go viral, but the one that converts is education, and education pairs well with either of the other 2. That sent me to tech reviewers, gamers and desk setup creators, whose audiences are already in a buying posture and whose format is functional by default. Engagement rose, people started reposting without being asked, and the sales showed up.
Insight
I stopped segmenting creators by reach and started segmenting by purchase intent. Reach is what you buy. Intent is what you get.
Fashion and lifestyle.Tech and product.
04
Ran the whole UGC program, all 7 steps
Pipeline
A template at every step. All 7 were built and run.
SourceFind creators by searching creator tags on Instagram. Hard cap of 30 to 40 DM requests a day, because past that the platform flags you as a spammer and the account is worth more than the volume.
Outreach1 opening message, 1 bump. Every creator logged in a database so nobody gets contacted twice by accident.
ScreenSignup form, a small creative task, then a 25 minute call on a booking link.
OnboardContract, shipping form, payment form. Then a 100% off code so the creator places their own order, which means address, phone model and product choice arrive as a clean order record instead of 5 DMs I have to transcribe.
TrainDiscord onboarding, brand and product knowledge, a content cheat sheet, an account warm up guide. Nobody films before that is done.
Run the week1 creator meeting: recap, shoutouts, performance reveal, next week's direction, feedback, metrics. Raw clips and posted videos land in Drive or it did not happen.
Pay on performanceSubmission form Sunday, uploads verified, accounts tracked, paid Monday. Base plus performance plus bonus tiers, so the ceiling belongs to the creator and the floor belongs to me.
Insight
The pipeline ran end to end, but not long enough to produce numbers worth quoting. We changed strategy partway through. What I can say is that the mechanism worked and the documents exist.
The harder thing is the constraint underneath it. A physical product has to fund a much longer loop than an app does. Pick the creator, check the product fit, collect shipping details, place the order, pack it, ship it, wait for it to land, wait for the unboxing. That is 2 to 3 weeks before a single frame gets shot, and every one of those steps is somewhere to lose a creator. You need budget and patience before the data means anything, and that is worth knowing before you start rather than after.
