Case studies / Advisory
Merchant Center setup · Creator ecommerceTikTok creates the demand. Google was there to catch it.
Kylee was already selling successfully on TikTok Live. She didn't need an agency to run ads, she needed Merchant Center and Shopping built properly so she could run them herself. Three months of setup, and the catalogue has since grown from a few hundred products to over 1,500 on the same foundation.
A TikTok Live seller, opening a second channel.
Kylee's Crystals Shop sells crystals and metaphysical goods, built first through live selling on TikTok, where the personality of the host and the moment of a stream do the selling.
We were brought in to build the Google Merchant Center and Shopping foundation so she could open a second, always-on channel and run it herself, without depending on an agency.
On TikTok she is the reason people buy. On Google she's invisible.
A live sale works because of the person hosting it: the personality, the moment a piece gets pulled, the scarcity of claiming it before someone else does. None of that exists in a Shopping result, where a listing competes on image, price, and relevance against sellers the buyer has never heard of. The two channels don't just look different. They work on opposite mechanics.
Demand created in the moment.
People arrive without intending to buy anything and leave with a piece they didn't know existed. The host, the pace, and the scarcity do the selling. Demand is generated, not captured, and most of it disappears when the stream ends.
- Buyer intent Created live
- What sells The host
- Timing Only while live
- Demand after Mostly uncaptured
Demand captured after the fact.
Someone who watched a live, didn't buy, and remembers the shop days later. Or someone searching for a specific stone with no idea who sells it. The listing has to do the work alone, which is exactly why the feed has to be right.
- Buyer intent Already there
- What sells The listing
- Timing Always on
- Demand after Captured
A working business with one channel.
Kylee's Crystals Shop was built on TikTok Live and was already selling well. The goal wasn't to fix anything, it was to open a second channel without depending on an agency to operate it.
Four things made the setup worth doing carefully.
Crystals break Shopping's assumptions
Most products have manufacturer barcodes. Crystals don't. Feeds that don't handle missing identifiers correctly get products disapproved, and the fix is in the setup rather than in the campaigns.
Hundreds of SKUs, with more coming
A catalogue this size and this varied can't be organised ad hoc. Whatever structure went in would have to survive the shop tripling in size, which it since has.
The language that sells is the language that flags
Crystals are described in terms of what they're believed to do. Merchant Center treats health and wellbeing claims carefully, so product data has to describe the object rather than promise an outcome.
She wanted to run it, not hand it over
A solo creator can't afford a permanent retainer and shouldn't need one for a Shopping account. Anything built for her rather than with her would have failed the brief.
Five parts, built once, handed over.
Each with the reasoning attached, so it could be extended without us.
Merchant Center from scratch
Account setup, verification, shipping, and returns configuration: the foundations that every later decision depends on and that are painful to correct once products are live on top of them.
A compliant account before a single product was submitted.
A feed built for a catalogue that would grow
Product data standards, categorisation, and identifier handling designed for hundreds of items with thousands in mind. Structuring a feed properly costs nothing extra at setup and saves a rebuild later. That's why the same structure still carries the shop at more than 1,500 products.
A foundation that scaled without being redone.
Product data that describes, not promises
Titles and descriptions written around what the product physically is: stone, size, form, finish, rather than what it's believed to do. It keeps listings eligible, and it happens to match what someone typing a specific stone into Google is actually looking for.
Listings that stay approved and match real searches.
Shopping structured so performance is readable
Product grouping set up so results can be read by type rather than as one blended number. On a catalogue this varied, a single campaign tells you that something sold, not what, or whether to buy more of it.
Decisions she could make from her own reporting.
Teach the running of it, not just the building
Training on how the account works, what to watch, and what to change when, so the handover was a capability rather than a login. That's the part that determines whether a setup engagement is still working a year later.
An owner who could run and extend it alone.
The catalogue tripled. The structure didn't change.
We're not publishing traffic, sales, or return figures, because we didn't run the account and don't hold them.
What we can point to is this: the shop went from a few hundred products to more than 1,500, and it's still running on the structure built during a three month engagement, managed entirely by Kylee.
That's the real test of a setup project. Plenty of accounts work on day one. Far fewer are still holding up after the catalogue has tripled and the person running them has never needed to call the people who built it.
It's also the honest recommendation for most creators in this position. If you already generate demand, you usually need a foundation and some training, not a retainer.
Four things for creators opening a second channel.
Catch the demand you already make
If people discover you live and buy later, brand search is the cheapest bridge between the two. Most creators never claim it, and someone else eventually will.
Structure the feed for the catalogue you'll have
Feed structure costs nothing extra at setup and is expensive to redo. Build for where the shop is going, not where it is.
Describe the object, not the outcome
In wellness adjacent categories, the language that sells is often the language that triggers disapprovals. Say what the thing is, let the buyer bring the meaning.
Buy a foundation, not a retainer
If you already have demand and can learn the tools, a short setup engagement usually beats paying someone monthly to run something you could run yourself.
What went into it.
Four of our services, delivered as setup and advisory.
Selling well on one channel and nowhere else?
If live selling or social is working for you, the demand you're creating is probably going uncaptured everywhere else. Tell us what you sell and we'll say whether you need a setup, a retainer, or neither.
No obligation · We'll tell you if you don't need us