Our Methodology

How We Evaluate Tea Products

Most tea recommendation sites give you a single “quality” rating and call it a day. We think that’s useless.

Here’s the problem: nearly every tea in our catalog is well-rated by customers. When everything scores 4.5 stars or above, a single quality number tells you nothing. A $9 pu-erh cake and a $60 box of Mighty Leaf tea bags can both be “excellent” — but they’re excellent for completely different people.

So instead of asking “how good is this tea?”, we ask four different questions.

The Four Dimensions

Quality & Craftsmanship

Is this tea made for someone who appreciates the craft?

We look for signals of depth: whether the tea rewards multiple steepings, whether it calls for traditional brewing methods like gongfu, whether it comes from a specific estate or mountain rather than a generic blend. We also look for complexity — teas described as layered, evolving, or nuanced score higher than those described as simple and straightforward.

A $9 pu-erh cake that rewards ten steeps in a gaiwan scores high here. A $60 box of convenient tea bags does not — and that’s fine. This isn’t a judgment of worth. It’s a signal for the right buyer.

Value

Is the price fair for what you’re getting?

We calculate value based on price per gram, but — and this is important — we don’t use a single yardstick for all teas. A pu-erh cake at $0.08 per gram is a perfectly normal, everyday purchase. Matcha at the same price would be suspiciously cheap. So our value assessment is category-aware: bulk teas like pu-erh, black tea, and chai are compared against bulk price ranges. Specialty teas like matcha are compared against specialty ranges.

Value scores are based on current Amazon pricing, not a snapshot from when we first listed the product. Prices change, and our assessments stay current.

Convenience

How easy is it to actually make a cup?

Some people want to weigh leaves on a gram scale and dial in water temperature. Others want to drop a bag in a mug and go. Neither is wrong — but they’re very different experiences.

We look at whether a tea requires special equipment, whether it’s forgiving of imprecise brewing, and whether it comes in a ready-to-use format like tea bags or sachets. A tea bag that brews well in any mug scores high. A delicate gyokuro that needs precise 60°C water and a kyusu scores low — which, again, isn’t a criticism. It’s information.

Gift Suitability

Would this make a good gift?

This one’s simpler. We look at packaging (tins, gift sets, premium presentation), origin specificity (a first-flush Darjeeling is a more interesting gift than a generic “black tea”), and whether the product is mentioned in the context of gifting by people who’ve bought it.

How These Dimensions Show Up

The four dimensions aren’t buried in a database somewhere — they’re how you navigate TeaDelight.

Browse by what matters to you. On any category or tag page, you can sort products by Quality, Value, Convenience, or Gift Suitability. Looking for the best-value green teas? Sort by Value. Want to find the most gift-worthy oolongs? Sort by Gift. The sorting controls are right there, above the product grid.

Every product card tells you what it’s best at. Instead of just showing a price and a star rating, each product highlights its strongest dimensions. A pu-erh cake might show “Connoisseur Pick” and “Sweet Spot” — telling you at a glance that it’s both a quality tea and fairly priced. A box of tea bags might show “No Fuss” and “Great Value” — different strengths for a different buyer.

Articles recommend as you read. When you’re reading an article about oolong tea, the product recommendations at the bottom know what section you’re in. You’ll see oolongs, not random black teas. And you can filter those recommendations by the same four dimensions.

Tea and accessories are handled differently. Value is calculated from price-per-gram, which doesn’t mean anything for teapots. So when you sort by Value, you’ll only see actual teas. But sort by Quality, and you might see a gaiwan alongside an oolong — because that’s a useful pairing.

What We Don’t Do

Being transparent about our limits matters as much as explaining our approach.

We removed metrics that didn’t work. Early on, we tried generating overall satisfaction scores, confidence ratings, and “beginner-friendly” tags. The satisfaction scores reflected how much feedback was available, not actual quality. The confidence scores had no correlation with reality. And the beginner-friendly tags? 96% of products received them — which makes the tag meaningless. We scrapped all three rather than pretend they were useful.

We acknowledge selection bias. Our catalog features products that are generally well-received. You won’t find a single 2-star tea here — not because we’re hiding them, but because we focus on products worth recommending. Every product in our catalog is rated 4 stars or above. This means our scoring differentiates between “good” and “great,” not between “bad” and “good.”

Small samples mean preliminary findings. Some products have extensive customer feedback to draw from. Others have very little. We’re upfront about this — a product with limited feedback gets treated as preliminary, not definitive.

Taste is subjective. We can tell you a tea is complex, earthy, and rewards patient brewing. We can’t tell you whether you’ll like it. Individual preferences, water quality, brewing habits — these all vary. Our scores help narrow the field, not make the decision for you.

Our Affiliate Relationship

We participate in the Amazon Associates program. When you click “Check Price on Amazon” and make a purchase, we earn a commission. This is how TeaDelight stays running.

Here’s what that doesn’t mean: our scoring is not influenced by commission rates. We don’t know — and don’t check — which products pay higher commissions. The scoring criteria are the same for every product: the signals we look for, the category-aware price comparisons, the customer feedback we draw from. A product that scores well on quality scores well because it shows signals of craftsmanship, not because it pays us more.

We think the best way to earn your trust is to help you find tea you’ll actually enjoy and come back for. A bad recommendation might earn a commission once. A good one earns a reader who returns.