We Don't Send You Coffee We Think Is Good. We Send You Coffee We Think You'll Like.
People ask me all the time: "What's the best coffee?"
My answer is always the same. It's the coffee you enjoy the most.
That might sound like a non-answer, but it's the foundation of everything we're building at Siip. And it's a fundamentally different starting point from how most of the coffee industry works.
The problem with "trust us, it's great"
Most coffee subscriptions work like this. They curate great coffees. They tell you to trust their taste. Whatever they send you is supposed to be good. And sometimes it is. Sometimes it's not your thing at all. But the experience stays roughly the same over time because the system never actually learns what you like. It just keeps sending what they think is good.
This is the sommelier model. Someone with expertise picks for the table. And that works fine when you're at a restaurant and the sommelier has five minutes to size you up. But for a product you drink every single day? You deserve better than someone else's best guess.
What we're building instead
At Siip, we inverted the model. We don't start with "what's the best coffee in our catalog?" We start with "what does this specific person like?"
Every coffee you rate, whether it's from your subscription, from a cafe, or from a bag you scanned while traveling, teaches our algorithm something about your palate. Not broad categories like "I like medium roast." Specific patterns. That you tend to prefer washed-process coffees from higher altitudes. That you rate natural Ethiopians higher than you'd expect from your stated roast preference. That there's something about Kenyan SL-28 variety coffees that consistently scores high for you.
These are patterns you might not even be aware of. But they're real, and they emerge from data. From the actual coffees you've rated, not from a quiz you took once.
The result: each month's delivery is better matched than the last. Your first bag might be an 85% match. By your fourth or fifth delivery, you're regularly above 93%. The subscription genuinely improves.
Why this matters more than you think
I've spent 20 years in coffee. Picking cherries on a farm in Costa Rica as a teenager. Sourcing with buyers from Blue Bottle. Building data systems at Enveritas. Watching a farming family I've known since I was 17 go from being rejected by Blue Bottle to placing 5th at Cup of Excellence. And the single biggest problem I've seen in this industry is that consumers don't have the tools to understand what they like.
It's not that people don't care about coffee. It's that the industry hasn't given them a framework to care effectively.
Think about wine for a second. When I go to a wine shop, I know I like Pinot Noir. I love Italian Barolos and Barbarescos from the Piemonte region. I know this because at some point I tried a wine, loved it, looked at the label, and learned what it was. The wine industry's data infrastructure made that possible. I didn't go to sommelier school. The structure was just there for me to learn from.
Coffee doesn't have this yet. You go to a specialty shop and the bag says "tasting notes: jasmine, bergamot, blackberry." Okay, but what does that mean for you? Should you buy it? Will you like it? There's no framework connecting those descriptors to your personal preferences.
That's what Siip is. We're building the data infrastructure that lets every coffee drinker, from the person who says "I just like strong coffee" to the specialty nerd debating processing methods, understand their own palate and find more of what they love.
You already know more than you think
One thing I've learned from putting coffee in front of people for two decades: consumers massively underestimate their own palate.
I've done this with friends who say they don't know anything about coffee. I'll put two coffees in front of them. They always have an opinion. Always. They can clearly say "I like this one better." And it's not always the one you'd expect. The person who says they only drink dark roast sometimes picks a medium-roast Brazilian. The person who says specialty coffee isn't for them rates a well-roasted single origin above their usual grocery store blend.
People aren't lacking in taste. They're lacking in vocabulary and framework. There's a massive difference between those two things.
What this means for your subscription
When you subscribe to Siip, here's what actually happens:
Month 1: You take a taste quiz. We match you from our database. It's a solid starting point, similar to what any subscription would do.
Month 2: You've rated your first 2 bags. The algorithm now has real data, not just quiz answers. Your match quality improves.
Month 3-4: You've rated 4-8 coffees, plus any you've scanned from cafes or stores. Patterns are emerging. The algorithm is discovering things about your palate that a quiz could never capture.
Month 5 and beyond: With 10+ ratings, your matches are consistently precise. You're receiving coffees you would never have found on your own, from origins you might never have tried, that hit your taste profile with remarkable accuracy. And you're far less likely to open a bag and think "this isn't for me."
That's what I mean when I say we don't send you coffee we think is good. We send you coffee we think you'll like. And we get better at it every single month.
Try the free Siip app (iOS/Android) to start building your taste profile. Scan and rate any coffee bag, anywhere in the world. When you're ready for a subscription that actually learns, your profile is already waiting.