Why Your Wine App Doesn't Actually Know Your Taste

Here is something nobody in the wine app world likes to talk about. The rating you see in most wine apps is not really about you. It never was.

It is about the 15,000 people who scanned that bottle before you.

• The professional who drinks 200 wines a month with a palate most people would not recognise as normal.
• The person who only drinks sweet whites and rated this dry red a 1 because they ordered it by mistake.
• The natural wine devotee who marks down anything that has seen oak.
• The hobbyist who gave it five stars in Tuscany, slightly sunburned, on their third glass.
• The person who genuinely loves this wine. The person who would never buy it again.

All of them, collapsed into the same number, handed to you as if it means something about your taste. It does not.

The problem with average ratings:

Average ratings are useful for some things. They tell you whether a wine is broadly liked and technically competent. If a wine has 2,000 ratings and averages 4.2 stars, it is probably a decent bottle.

But that is not the same as knowing whether you will love it.

Your palate is yours alone, shaped by biology, memory, and habit, and always evolving. The way you preference for tannins, your tolerance for acidity, your lean toward fruit-forward or earthy profiles: all of it shifts over time, and varies with context.

A wine you rate 3 stars on a Tuesday night might be a 5-star experience for someone else at a dinner party in the south of France.

So when a wine app tells you a bottle is a "good match”, it is worth asking: a match for whom, exactly? If that score comes from community ratings, even weighted by an algorithm, it is still based on other people's palates. Not yours.

"Personalised" usually isn't…

The word “personalisation” gets thrown around a lot in wine apps. Most of the time it means: we took the crowd data and filtered it based on a few preferences you told us about in a setup screen.

It’s a start. But it is not personalisation.

Real personalisation means the app learns from your behaviour. Your ratings, your cellar, the wines you return to across regions and varietals over time. Gioia takes all of it and builds a picture of who you are as a wine drinker.

Like fingerprints, that picture is unique to every single user.

How Gioia actually learns your taste:

When we built Gioia, this was the problem we kept coming back to. The existing tools were great at telling you what wine is popular. None of them were built to tell you what wine is right for you, specifically.

So we formulated individual user profiles and match scores from scratch to facilitate better personal recommendations.

Every wine you rate feeds your taste profile. The more you use the app, the more accurate Gioia understands your palate. The more accurately it understands your palate, the better the match scores and recommendations get.

When Gioia scores a new wine against your profile, the result is a match percentage calculated entirely for you, with no one else’s opinion in the mix.

Two people. Same bottle. Different answers.

Consider two people standing in the same wine shop, reaching for the same bottle.

Ava is a seasoned Burgundy drinker. She spent years working through the villages of the Côte de Nuits. She has a cellar full of Pinot Noir. She loves wines that are light on their feet, high in acidity, earthy and restrained — the kind that make you think rather than just drink. She rarely finishes a glass of anything over 14.5%.

Javier got seriously into wine a couple of years ago and discovered he loves bold, generous reds. Malbec from Mendoza, Shiraz from the Barossa, Napa Cabernet Sauvignon and Franc with ripe dark fruit and toasted oak. The bigger the better. He has never knowingly enjoyed a wine described as "delicate".

Both scan the same bottle. A highly extracted, full-bodied Californian Cabernet Sauvignon. Well made. Broadly popular.

A crowd-based app, like Vivino, shows both of them the same score. 90%. The implicit message: this wine is good, you will probably like it.

Gioia shows her 29% and him 94%.

Not because the wine changed between the two scans. Because one of them is going to love it and one is going to pour half of it down the sink — and a good recommendation should know the difference.

A 29% match doesnt mean the wine is bad, but based on everything you have shown us about your taste, this is probably not the best bottle for Ava. She should try the Chambolle-Musigny two shelves over, an 89% match.

That is what a personal recommendation looks like with Gioia.

A small experiment worth trying:

Most people who get into wine go through a phase of chasing high scores and popular bottles. It works well enough when you are still figuring out what you like.

But at some point, a 4.4-star wine does nothing for you, and an obscure bottle you stumbled across is the best thing you have had all year. That is your palate asserting itself.

Next time you choose a wine, note the community rating. Then note how you actually feel about it after drinking it. Do this a few times.

You will quickly see the gap between what the crowd says and what your palate says. That gap is exactly what Gioia is designed to close.

Gioia is available on iOS and Android. Start building your palate profile here.