How do you decide which wines go on your shelf? If you are honest, for most wine shop owners it goes something like this: the sales rep visits, you taste a few bottles, you like three of them, you order. Maybe you look at what sold well last month. Maybe you have a feeling that rose will pick up now that the weather is warming. But it is largely intuition, experience, and personal taste.
There is nothing wrong with that -- that intuition is valuable. But it is not enough. Online wine retailers continuously optimize their assortment based on data: sales trends, search behavior, seasonal patterns, click rates. They know exactly what their customers want. The physical wine shop that adopts the same data-driven approach -- combined with the human intuition that online lacks -- has an unbeatable advantage.
Intuition is a valuable starting point, but it has three systematic weaknesses:
Wine shop owners are wine lovers. That is their strength, but also their pitfall. If you love Barolo, chances are your assortment is stronger in Italian reds than your customer base justifies. You buy what you personally enjoy, not what your customers buy most.
This is not a theoretical problem. Analysis of wine shop assortments shows that most shops dedicate 20-30% of their shelf space to wines that generate less than 5% of revenue. That is dead capital on a shelf that needs to earn money.
Wine sales reps are excellent sellers. They come with enthusiastic stories, attractive introductory prices, and sample bottles that are hard to refuse. But their interest is not your interest: they want to sell their portfolio, you want to serve your customers.
That does not mean you should ignore reps -- their offerings can be excellent. But the decision about what to buy should rest with you, supported by data, not with the rep who has the best pitch.
Without data you miss patterns that are very real. Rose does not sell "when it gets warm" -- rose starts moving in the first sunny week of March and peaks in June. Full-bodied red stew wines do not peak in "autumn" but specifically in the weeks around the first cold evenings, usually late October. That timing makes the difference between a full shelf and an empty shelf at the right moment.
You do not need to be a data scientist. Most data sources you already have, you just are not using them systematically:
Your register contains gold. Export your sales data monthly (most modern POS systems can do this) and answer four questions:
Every time a customer asks for something you do not carry, that is a data point. "Do you have natural wine?" "I am looking for a good Albarino." "Anything from the Jura?" Most shops let this information evaporate.
Solution: keep a notebook at the register (or a note app on your phone) and write down every request. After a month you have a list of concrete gaps in your assortment -- directly from your customers.
If you have QR codes on the shelf (see our guide to QR codes), you have a rich data source. Scan data tells you:
Platforms like SommelierX provide insight into which food pairings are most popular in your region. If you know that "wine with pasta" and "wine with cheese" are the two most searched pairings among your customers, you know your Italian assortment and your wine-with-cheese selection need to be strong.
This is data you did not have before. The combination of what customers are cooking and which wine they are looking for is the most direct indicator of what you should stock.
A data-driven assortment is not static. It consists of three layers:
These are your proven sellers. The wines that sell month after month and that your regular customers expect to find. This layer is stable: you only change it when a wine goes out of production or structurally underperforms.
Characteristics of core wines:
These are wines that change per season or quarter. They keep the assortment fresh and give customers a reason to come back. "What do you have that is new?" is a question you want to hear.
Rotation triggers:
These are your surprise picks. Wines that you personally find fascinating, that your customers do not know, and that strengthen your story as a shop. A Greek Xinomavro, a Hungarian Furmint, a South African Chenin Blanc from an unknown producer.
The experimental layer is where your personal taste and expertise shine. It is also where you differentiate from the supermarket and the webshop. But limit it to 5-10% -- otherwise your assortment becomes a museum instead of a shop.
Collecting data is step 1. Using that data is step 2. Schedule one hour every month for an assortment review. Here is the format:
A good assortment is not just the right wines -- it is also the right price spread. The most effective price matrix for a wine shop:
Compare this distribution with your actual sales data. If 50% of your revenue comes from the 10-18 euro category but only 30% of your shelf is dedicated to it, you are leaving money on the table.
Based on sales data from hundreds of wine shops, this is the seasonal pattern most shops see:
The most valuable data for assortment decisions is what your customers want, not what they buy. Because what they buy is limited by what you offer. What they want reveals where the demand is -- even if you do not have supply for it yet.
The SommelierX retailer dashboard shows exactly that: which food pairings your customers search for, which wine styles are popular in your region, and which categories are growing. If you see that "wine with Asian food" is the fastest-growing search category in your shop, you know you need to expand your Asian-compatible wines (Riesling, Gewurztraminer, off-dry styles).
That data is the difference between reactive purchasing (what sells well?) and proactive purchasing (what is going to sell well?). And proactive always wins.
Combine four data sources: POS sales data, customer requests, seasonal trends, and pairing popularity. Data does not replace your wine expertise but supports your decisions with facts instead of feeling alone.
Quality over quantity. A well-curated assortment of 200-400 wines outsells an unstructured assortment of 800. The key is breadth across categories combined with depth in your strongest niches.
Keep a core of 60-70% proven sellers that remain stable. Rotate 30-40% per season or quarter. Evaluate monthly which wines have been sitting too long and replace those first.
The SommelierX retailer dashboard shows which pairings and wine styles are popular in your shop. Data-driven purchasing starts here.
Explore options for shopsMore for wine businesses: View all B2B articles