An influencer’s apparel brand, launching on limited drops, needed a store that would earn trust on its own, not lean on her name to carry every sale. We designed it from a library of 800+ e-commerce research studies.
| Client | Influencer-led apparel brand, EU |
| Sector | E-commerce – fashion, limited-drop model |
| MARKET | EU |
| Engagement | Evidence-led store and product-page design |
| METHOD | Applied research — 800+ UX and e-commerce studies — not on-site behavioural testing |
| Constraint | Scarcity-based inventory · a founder whose fame was an asset and a trap |
Two ways to design a store
There are only two.
You can copy a competitor, follow your taste, argue it out in a meeting and then test your guesses on live traffic, paying for the wrong ones in lost sales until the data corrects you. That’s conversion optimisation, and it’s the right tool when a site is already running and you can watch real users struggle.
This wasn’t that. This was a brand getting its store built. There were no users to watch yet, and no traffic to spend learning on. So we used the other method: design it correctly from the start, using what is already known.
Most stores are still built on instinct, and instinct quietly reproduces the same avoidable mistakes. Our practice is built on the opposite premise. We maintain a library of 800+ e-commerce and UX research studies, and we design from it. Every decision below traces to evidence and not to a hunch we’d have to validate later at the client’s expense.
The gap between obvious and actual
Here is the uncomfortable thing about e-commerce best practice: most of it sounds obvious, and most stores still get it wrong.
We have audited hundreds of shops. The number that are missing the basics is not small. “Obvious” and “present” turn out to be almost unrelated.
That gap is the entire opportunity. You don’t need an exotic growth hack to beat the field. You need to actually implement the things everyone already agrees with and almost nobody does. This store was built to be on the right side of that gap on every point that the research says matters.
The specific studies and the reasoning behind each choice on our research platform.
Solution 1: Design for scarcity without wasting it
The brand runs on limited drops. When a piece sells out, it’s gone until the next batch.
Handled carelessly, scarcity is pure waste. A sold-out product is a visitor who arrived wanting to buy (the most valuable person on the site) meeting a dead end and leaving with nowhere to put that intent. The demand was real and you let it evaporate.
So we built the sold-out state to capture intent rather than lose it. A visitor who reaches an unavailable item can leave an email to be notified, and the store signals roughly when the next batch lands. The person most ready to buy is no longer a bounce. They’re a known lead and a head start on the next drop. The anticipation that scarcity creates works for the brand instead of against it.
There’s a well-documented behavioural reason scarcity drives demand at all. There’s an equally documented reason it backfires when it produces only frustration. We designed for the first and against the second.
Solution 2: Take the risk out of buying a hoodie you can’t touch
The single biggest reason a first-time apparel buyer hesitates and the single biggest driver of returns is uncertainty about fit. Online, they can’t try it on. So the store has to do the reassuring, and most stores do it badly or not at all.
We layered the product page to remove that uncertainty in stages:
- A size table in centimetres – actual garment measurements, not a vague S/M/L that means something different in every brand.
- Instructions for how to measure – placed directly beneath the table, because a size chart only helps a buyer who knows what to measure against. The table answers how big is it; the instructions answer how big am I. One without the other is half a solution, and the research is clear that the gap is where the wrong-size order is born.
- Care information – how to look after the garment, addressed before purchase rather than discovered after a wash goes wrong.
None of this is exotic. All of it is evidence-backed, and a startling number of stores ship without it. Every layer removes a specific reason to hesitate and a specific reason to return.
Solution 3: Let the founder support the brand, not become it
The brand belongs to an influencer with a real audience. The tempting play – the one most creator brands make – is to point the entire store at her: her face on everything, buy it because it’s mine. It converts her existing fans on day one.
A store built entirely on a founder’s fame sells to her followers and stops. It reads as merch, not as a label and merch has a short life and a low ceiling, because it asks the customer to buy a relationship, not a product. The moment you want a customer who doesn’t already follow her, you have nothing to offer them, because you built the whole store on a connection they don’t have.
So we made a deliberate choice: her presence runs through the store as support, not foundation. It’s there – enough to reassure her audience the brand is authentically hers and to convert them easily. But the store stands on its own as a credible apparel label: the product, the fit, the quality, the experience carry the sale. A stranger who has never heard of her can land on the store and buy, because nothing on the critical path requires them to care who she is.
That’s the difference between a store that converts her current followers and a brand that can eventually outgrow her. We built the second one.
Why design from research instead of testing your way there
A fair question: if conversion testing works, why not just launch something and optimise?
Because testing has a price, and it’s paid in live traffic and lost sales. Every wrong guess you put in front of real buyers costs conversions until the data catches it. On a new store – especially one running scarce inventory, where every drop is a limited number of chances – you cannot afford to learn the expensive way on the basics. There aren’t enough sales to burn.
Testing is the right tool for the questions research can’t answer – the ones specific to your users, your product, your particular funnel. It is the wrong tool for questions that thousands of studies have already answered. Using an A/B test to rediscover that fit uncertainty drives returns is like running an experiment to check whether water is wet. Spend your traffic on what’s genuinely unknown. Get the known things right on the first build.
That’s the discipline: research decides what’s already settled; testing decides what’s still open.
If you’re in the same position
Someone has already run your experiment. Before you test, argue, or guess, ask whether the answer already exists. For most foundational e-commerce decisions, it does – measured across thousands of stores.
Obvious and implemented are not the same thing. The elements that most reliably convert are the ones everyone nods along to and most stores still skip.
A creator brand that leans entirely on the creator has a ceiling. Use the founder’s audience to launch. Don’t build the store so that it can only ever sell to them. Support, not foundation.
The gap between what’s known and what’s done is where most of the money is. You rarely need a new idea. You need to actually do the things that are already proven which, going by the state of most stores, is rarer and more valuable than any growth hack.