A DACH wine retailer’s catalogue was written for experts. But its customers weren’t experts. We rebuilt the buying decision around food instead of grapes and the year after launch, sales rose 70%.
| Client | DACH-based online wine retailer |
| Sector | E-commerce – food & beverage |
| Market | German-speaking Europe |
| Engagement | Customer research + visual identity + webshop design + asset system |
| Constraint | Buyers wanted to buy wine confidently. The shop assumed they already could. |
| Timeline | 3 months |
| Delivered | 10 pages · food-pairing graphic system · tasting profile for every SKU · 2 design iterations |
| Outcome | +70% sales in the first year after launch · corroborated by a post-purchase survey of 154 buyers |
The problem
We started with customer research, and it surfaced something the client’s analytics could never have shown.
People wanted to bring wine to a dinner, to a host, to a gift exchange. They wanted to know wine and they had nowhere to learn it. So, at the moment of purchase, they froze.
Not because the selection was bad. Because they had no idea which bottle went with what they were eating, and no way to find out without admitting they didn’t know.
This is the defining condition of the category, and almost every wine shop online gets it backwards. The catalogue lists grape, region, vintage, alcohol content, and a sommelier’s tasting note written in sommelier language. It is a beautifully organised reference document for people who already know how to read it.
Which means the shop is fully optimised for its smallest segment and quietly abandoning the majority. Those buyers do one of three things: default to the label they recognise, default to the cheapest, or leave. All three cap average order value. None of them come back better informed.
The client wasn’t losing sales to a competitor. They were losing them to their customer’s fear of looking stupid.
Why the obvious answer fails
Every solution the category reaches for makes the same mistake – it assumes the buyer has a vocabulary they don’t have.
More filters. Filter by grape, region, body, price. This is a tool for someone who can already specify what they want. Giving a more precise instrument to a person who doesn’t know what they’re looking for does nothing except make the uncertainty more granular.
Better expert copy. More sommelier notes, longer descriptions, richer language. Gooseberry, flinty minerality, a long finish. This only confirms the buyer’s anxiety. Every line of expert copy tells the non-expert, precisely and politely, that they are an outsider here.
A recommendation quiz. Answers the question, teaches nothing. The buyer converts once and returns exactly as ignorant as before – dependent on the quiz forever, and no more confident about the thing they said they wanted to be confident about.
Bestseller badges. Herds every uncertain buyer onto the same ten SKUs, kills the long tail, and guarantees your margin lives on whatever is already popular.
Each of these tries to help the buyer choose. None of them makes the buyer competent. And competence was the thing they had told us, unprompted, that they wanted.
Solution 1: Start from what they already know
The buyer doesn’t know wine, but they know exactly what they are having for dinner.
So we anchored the entire decision in the knowledge they already possessed. Every wine received a food-pairing graphic – a clear visual system showing which dishes it belongs with: seafood, meats, cheeses, desserts.
Recognition instead of recall. Asking “which wine would you like?” is a recall task, and recall from an empty vocabulary is impossible. Showing “you’re serving fish – these three” is a recognition task. Recognition is dramatically cheaper, and it’s the difference between a decision the buyer can make and one they can’t.
Picture superiority (Paivio, 1971). Visual information is encoded and retained better than the same information in text. A pairing icon isn’t decoration standing in for the description. It outperforms the description, and it survives being skimmed.
The buyer no longer had to become a wine expert to buy wine. They only had to know what was on the plate.
Solution 2: Sell the sentence, not just the bottle
The second research finding was the one that changed the project.
Buyers didn’t only want to choose correctly. They wanted to arrive with the bottle and be able to say something about it. Wine bought for other people is a social purchase, and the job it is hired to do is not “taste good.” It is “make me look like someone who knows what they’re doing.”
So every wine got a tasting profile – its flavour characteristics laid out visually, plainly, in language a person could actually repeat out loud.
Social currency (Berger, 2013). People choose and share things that make them look good to others. A product that hands you something to say is doing more work than a product that merely satisfies you. The tasting profile wasn’t a specification. It was a script.
And it compounds. A buyer who learns something on a product page becomes a slightly more confident buyer next time, which means they browse wider, trade up, and stop defaulting to the one label they recognize.
What we built
- Complete visual identity and digital presence
- Food-pairing graphic system across the catalogue
- Tasting profile visual for every wine in the range
- One asset system, deployed everywhere: the same graphics ship to the webshop, paid ads, and social with no re-production.
- The logic extended offline. The short pairing descriptions went onto the company’s business cards. A business card that is useful to the person holding it does not get thrown away. Same principle as the whole project: give the customer something they can use, and they keep it.
The result
In the first twelve months after launch, sales rose 70% against the twelve months preceding it.
A post-purchase survey of 154 buyers found the pairing and tasting visuals to be a major factor in their decision to buy.
Self-reported attribution has its limits; people are unreliable narrators of their own choices. But consider the chain. The qualitative research predicted a specific mechanism – buyers are blocked by not knowing, and by the fear of appearing not to know. The features built on that mechanism shipped. Sales moved 70%. And buyers then independently named those features as the reason.
If you’re in the same position
You are, if you sell anything with an expert layer: wine, coffee, olive oil, skincare, supplements, audio, tools, components. Anything bought as a gift. Anything where the buyer suspects there is a right answer and fears they don’t know it.
Your catalogue answers the wrong question. It answers what is this? Your customer is asking is this right for me, and will I look foolish? Those are not the same question, and no amount of product data answers the second one.
Anchor the decision in knowledge the buyer already has. They don’t know grapes. They know what’s for dinner. Find the thing your customer is already an expert in and build the choice around that.
Design the asset once, deploy it everywhere. If a graphic works on the product page but has to be rebuilt for ads and rebuilt again for social, it was designed wrong.
And the one worth remembering: in any category with an expert layer, the product that wins isn’t the best one. It’s the one the buyer can defend. Give them the sentence, and you have not just made a sale, you have made someone who is willing to come back and buy something harder.