Why BeanScore exists

There has probably never been a better time to be into specialty coffee.

Independent roasters are sourcing incredible coffees from around the world. New origins, varieties, processing methods and flavour profiles appear constantly.

There is no shortage of good coffee.

And yet deciding what to buy next still feels harder than it should.

That's the problem BeanScore was built to solve.

But it didn't start as a coffee discovery app.

It started as something I wanted for myself.

It started with a Notes app

I was drinking more specialty coffee and wanted a better way of keeping track of what I tried.

Like a lot of people, my system wasn't particularly sophisticated.

Notes on my phone.

Photos of bags.

Trying to remember the name of a coffee I'd loved months earlier.

Occasionally scrolling back through old screenshots because I vaguely remembered that one Ethiopian coffee was really good.

I could remember whether I liked an individual coffee.

What I couldn't see was the bigger picture.

Why did I like it?

Was there something about the origin? The roast? The acidity? The flavours?

Were there patterns across the coffees I kept rating highly?

So I started building a more structured way of recording what I drank.

And then something interesting happened.

The patterns became obvious.

One of the first things I noticed was how consistently I enjoyed Ethiopian coffees.

It sounds almost laughably simple now.

But until I could see my drinking history together, I hadn't consciously realised it.

I'd spent years judging individual bags of coffee.

Suddenly, I was beginning to understand my taste.

That was the point where the idea changed. It wasn't really about tracking coffee anymore. It was about understanding taste.

That idea eventually became BeanScore.

"What's the best coffee?" is the wrong question

Coffee recommendations tend to revolve around a deceptively simple question:

What's the best coffee?

Ratings reinforce that.

One coffee has 4.8 stars.

Another has 4.2.

So presumably the first one is better.

Except taste doesn't really work like that.

Someone else's five-star coffee might be far more acidic than you enjoy.

A coffee with huge fermented fruit flavours might be somebody's dream cup and somebody else's nightmare.

You might consistently love coffees other people describe as safe, traditional or unexciting.

None of those people are wrong.

They simply have different tastes.

That's why the question I'm much more interested in is:

What's the best coffee for me?

A community rating can tell you whether people generally think a coffee is good.

It can't necessarily tell you whether you are going to enjoy drinking it.

BeanScore is designed around that distinction.

From ratings to Taste DNA

A single rating doesn't tell us very much.

But dozens of ratings start to reveal patterns.

That's the thinking behind Taste DNA.

As you rate coffees, BeanScore begins building a picture of the things you tend to enjoy.

That can include characteristics such as:

• acidity
• sweetness
• body
• origins you repeatedly rate highly
• roast styles you prefer
• flavours that appear in your favourite coffees
• brewing preferences

Instead of every coffee existing as an isolated five-star score, your ratings start becoming part of a wider picture of your palate.

The more you rate, the more useful that picture becomes.

And that creates a different way to approach discovery.

Instead of:

"People like this coffee."

BeanScore can work towards:

"Based on the coffees you've enjoyed, this looks like something you might like."

It's a subtle difference, but I think it's an important one.

Discovery should be personal

The goal isn't to tell everybody to drink the same highly rated coffees.

That would be boring.

It's to help someone who loves delicate, floral washed coffees discover more of them.

Or help someone who prefers chocolate-heavy, lower-acidity coffees avoid repeatedly buying bags that were never likely to suit them.

Or help somebody realise — like I did — that there is a pattern running through the coffees they already love.

Coffee discovery becomes significantly more useful when the starting point is the person drinking it.

But there's another problem.

Even the best recommendation isn't particularly useful if you can't actually buy the coffee.

Coffee doesn't stand still

Specialty coffee changes quickly.

Harvests change.

Lots sell out.

Roasters rotate their ranges.

New coffees arrive and old coffees disappear.

The same farm can produce different lots. A coffee may return the following year with a different crop, process or flavour profile.

Coffee is inherently seasonal.

And that creates a problem for traditional coffee databases.

You can build a catalogue containing an enormous number of coffees, but size alone doesn't make that catalogue useful.

A catalogue containing 10,000 coffees isn't very useful if the coffee you want to try disappeared two years ago.

BeanScore isn't intended to become a museum of every coffee that has ever existed.

Discovery needs a relevance layer.

If BeanScore recommends something to you, the aim is that it's a coffee that's actually available — something you can discover today and potentially order today.

That means treating coffee as something alive, rather than a static product database.

It's an idea I'll explore properly in another Journal entry because it has become one of the principles behind how BeanScore is being built.

More than a coffee tracker

BeanScore still lets me do the thing I originally wanted.

Remember the coffees I've drunk.

But the bigger idea now goes much further than that.

It's about connecting three things:

Rate what you've tried.

Understand what you like.

Discover what to try next.

Your past coffee becomes useful information for your next one.

And over time, you start building something that most of us have without ever really being able to see:

your own coffee taste profile.

What the BeanScore Journal is for

This Journal won't simply be a changelog for the app.

BeanScore gives me an excuse to spend an unreasonable amount of time thinking about coffee, taste and the way people discover what they drink.

So this is where I'll share some of that.

There'll be articles about coffee discovery and taste — ratings, tasting notes, preferences and why different people experience the same coffee differently.

There'll be coffee education — origins, processing, roast levels, flavour, brewing and some of the terminology surrounding specialty coffee.

As BeanScore grows, I'll also share findings from the platform where there's something genuinely interesting in the data.

And occasionally I'll write about building BeanScore itself — the product decisions, ideas and problems behind trying to build a better coffee discovery system.

Some articles will have almost nothing to do with the app.

That's intentional.

The Journal should be useful to somebody who's interested in specialty coffee even if they've never used BeanScore.

This is only the start

BeanScore is now live, starting in the UK.

If you've already got a mental list of coffees you loved, photos of forgotten bags in your camera roll, or a Notes app full of half-remembered tasting notes, you probably understand why I started building it.

Now I want to see how much further the idea can go.

Or, if you'd like to know more about the story behind the app, read about BeanScore →

Jack

Founder, BeanScore