Search for profitable products to sell online and you will find lists. Lists are cheap to write and almost useless, because profitability is not a property of a product — it is a property of a product plus a shipping weight, a shelf life, a return rate, a regulatory file, a reorder curve and a minimum order quantity you have to pay for up front. Two sellers can buy the identical herbal tea from the identical producer and one makes money while the other quietly runs out of cash. The difference is almost never the product idea.
So this guide does it in the other order. First the variables that actually decide whether a food or wellness product makes money online. Then a scoring table applying those variables to five concrete categories we know well — herbal and fruit tea, dried fruit, snack mixes, fruit powders and essential oils. Then the situations where the honest answer is: do not launch this.
One warning before we start, because it changes what you can even say in your listing.
The claim you are not allowed to make
A large share of food and wellness marketing you see online is illegal. In the EU, nutrition and health claims on foods are governed by Regulation (EC) No 1924/2006, and the rule is restrictive: a health claim may only be made if it appears on the EU register of authorised claims, in the authorised wording, for that substance. Almost no botanical infusion carries an authorised health claim. That means "supports digestion", "boosts immunity", "helps you sleep", "detox" and every variant of them are not marketing choices you get to make — they are enforcement exposure, and marketplaces increasingly remove listings that use them.
Why does this belong in a profitability guide? Because it removes a lever most beginners assume they have. You cannot rescue a weak product with a strong benefit story, because you are not permitted to tell one. What is left is sensory description (aroma, infusion colour, taste, texture), origin, quality documentation, format convenience and packaging. Products that can win on those attributes are viable online. Products that only sell when someone promises a physiological outcome are not a business — they are a takedown waiting to happen. Plan your category on that basis.
The six variables that decide margin
1. Value density: price per kilogram against shipping weight
This is the single most predictive variable in food e-commerce, and the one beginners miss. What matters is not your unit price but your unit price divided by the parcel weight you have to move. A 40 g box of tea bags and a 1 kg bag of dried apricots may retail within the same range, but one ships in a letterbox-friendly parcel and the other does not, and the difference recurs on every single order and every single return.
Value density also decides how far you can ship your inbound stock. A high-value-density product tolerates air freight in an emergency; a low one is locked to sea or road, which locks your reorder cycle to a longer horizon. Our Incoterms guide for natural products covers how inbound cost gets allocated, and it matters most exactly where value density is weakest.
2. Shelf life
Shelf life sets the maximum size of a batch you can responsibly buy. A product with a long, stable shelf life lets you order a larger, cheaper run and sell it over a year. A short shelf life forces small, frequent, more expensive runs — and every unsold unit at expiry is a total write-off, not a discount opportunity. When you model margin, model it after write-offs, not before. A category with a healthy gross margin and a meaningful expiry loss rate can be worse than a thinner-margin category with none.
Shelf life also interacts with marketplace fulfilment. Fulfilment services typically require a minimum remaining shelf life on receipt and dispose of stock that falls below it, so your effective sellable window is shorter than the printed one. Check the current policy of whichever platform you use — these rules change and they differ by marketplace and by country.
3. Return, breakage and damage rate
Food returns are not like electronics returns: an opened food item usually cannot be resold, so a return is close to a full loss plus two shipping legs. Breakage is separate and often larger. Glass bottles break. Powders leak from badly sealed pouches. Chocolate-coated anything arrives melted in July. Fruit that was dried to a soft texture compacts in transit and looks worse than it is, which generates "not as described" claims that are really packaging failures.
The fix is usually packaging specification rather than product change, which is why format decisions belong at sourcing stage and not after the first hundred complaints. Our doypack retail packaging guide covers the pouch specifications — barrier, seal, zip, gusset — that determine whether a dried fruit line arrives intact.
4. Regulatory load
Every food product carries a compliance file: mandatory labelling under EU food information rules, allergen declaration, ingredient list, net quantity, date marking, the name and address of the food business operator, lot marking and traceability. Some categories add more — sulphite declaration on many dried fruits, specific rules for products that sit near the cosmetic or supplement boundary, and separate regimes entirely for products classified outside food.
Regulatory load is a real cost, and it is mostly fixed rather than variable, which means it punishes small ranges and rewards focus. Ten SKUs across four regulatory regimes is a part-time compliance job; three SKUs in one regime is a checklist. Our notes on sulphites and sorbates labelling for dried fruit and EU lot tracking and traceability show what that file actually contains.
5. Repeat purchase rate
Acquisition cost is paid once; margin is earned every time. A consumable that a customer finishes in three to six weeks and rebuys can absorb an acquisition cost that would destroy a one-off purchase. This is the reason so many durable e-commerce businesses are built on consumables — not because the products are exciting, but because the second order carries no acquisition cost.
Judge a category by how naturally it repeats without you doing anything clever. Daily-use consumables repeat. Gift formats do not, or repeat annually at best. Novelty flavours repeat once and stop. If your plan depends on inventing a subscription for a product nobody consumes on a schedule, the repeat rate is not real.
6. Capital tied up in the MOQ
The last variable is the one that kills otherwise sound plans. Your minimum order quantity, multiplied by unit cost, multiplied by the number of SKUs, is money you cannot spend on advertising, and you get it back only at the pace of sell-through. Branded stock is worse than generic stock here, because nobody else can sell a box with your logo on it.
This is why we treat entry MOQ as a margin variable and not an operational footnote. A category where you can enter with a genuinely small run lets you learn from real sales before committing capital. Our MOQ and lead time guide explains why the number is what it is on a small-batch line, and the wholesale versus private label comparison covers the capital difference between the two models.
Scoring five categories against the six variables
The table below is our qualitative read, based on what we manufacture and ship. It is a starting point for your own analysis, not a verdict — your channel, country and packaging choices move several of these ratings.
| Category | Value density | Shelf life | Breakage / return risk | Regulatory load | Repeat rate | Capital tied at entry MOQ |
|---|---|---|---|---|---|---|
| Herbal and fruit tea, bagged | High — light unit, meaningful retail price | Long and stable when packed with a moisture barrier | Low; crush risk is a carton-strength problem, not a product problem | Moderate — food labelling, allergen and botanical rules; claims heavily restricted | High — a daily consumable finished in weeks | Low to moderate; small runs are achievable per SKU |
| Dried fruit, retail packs | Low to moderate — heavy relative to price | Long, but texture and colour drift before the date | Moderate — compaction, sugaring and "not as described" claims | Higher — sulphite declaration, mycotoxin and residue limits, origin | Moderate — repeats as snacking, not daily | Moderate; run sizes agreed per fruit and format |
| Snack and trail mixes | Low — heaviest of the five per euro | Set by the shortest-lived component in the blend | Moderate to high — separation, oil migration, melt risk with coated pieces | Higher — multi-ingredient allergen matrix, several origins to document | Moderate to high | Moderate to high; multiple raw materials to buy at once |
| Fruit powders | High — light, concentrated, high price per kilogram | Long if moisture and oxygen are controlled; poor if not | Moderate — clumping, leakage, spill claims | Moderate to high — depends heavily on how the product is positioned | Moderate — depends on whether it enters a daily routine | Moderate; processing minimums apply |
| Essential oils | Very high — small volume, high price per kilogram | Long, but oxidation-sensitive and batch-variable | High — glass breakage, leakage, transport restrictions for some oils | High — outside food labelling entirely; separate regime and documentation | Low to moderate — slow-consumption product | Moderate; small volumes but higher unit values |
Read the table as a set of trade-offs rather than a ranking. Essential oils have superb value density and terrible breakage economics. Dried fruit has honest, broad demand and the worst weight profile. Bagged tea is the most balanced of the five, which is the unglamorous reason it keeps appearing in successful small e-commerce ranges — and, in fairness, the reason it is also crowded.
An illustrative unit walk-through
The following numbers are illustrative only. They are a worked structure, not our prices and not a forecast. Your costs will differ by product, format, volume, destination and channel, and you must substitute your own quoted figures before drawing any conclusion.
Take a hypothetical retail box of tea bags. Build the cost stack in this order: manufactured unit cost (raw material, bagging, envelope or tag, printed box) → inbound freight and duty per unit → inbound handling and storage per unit → outbound pick, pack and delivery per unit → channel commission and payment fees at whatever rates currently apply in your category and country → expected returns, breakage and expiry write-off per unit → advertising cost per acquired order, divided by expected orders per customer.
Two observations that hold regardless of the numbers you insert. First, the last two lines — write-off and acquisition — are where most first-time models go wrong, because they are the two that get left out. Second, the cost stack is per unit but the MOQ is a lump sum, so a plan can be margin-positive and still fail on cash timing. Model both, and model them monthly.
If a product only works at a large run size, treat that as a warning rather than an opportunity. Prove demand at the smallest viable run, then take the unit-cost improvement when you have sales data to justify it. Our e-commerce launch roadmap sequences that properly.
When the honest answer is "do not launch this"
Some things are worth saying plainly.
Do not launch a category whose selling story requires a health, slimming or detox angle. You cannot legally tell that story in the EU, and a product that has no other reason to exist has no reason to exist.
Do not launch ten SKUs. The regulatory file, the artwork, the photography, the stock and the attention all multiply, and none of the fixed costs get shared as efficiently as you imagine. Two to four SKUs is a range; ten is a warehouse.
Do not launch a low-value-density product into a market you have to ship long distances into with no local stock. The freight will eat the category. If you must, plan a stock point closer to the customer — for us that is our Solingen, Germany warehouse, which turns deliveries into intra-EU movements.
Do not launch on a specification you have not tasted, weighed and photographed yourself. Order samples, evaluate them against a written checklist, and treat the sample stage as data collection rather than a formality; our B2B sample order guide sets out how.
And do not assume any marketplace rule you read — including in this article — is still current. Fulfilment shelf-life minimums, food category approvals, prep and labelling requirements and commission structures all change. Check the platform's current policy for your country before you build a plan around it.
A short pre-commitment checklist
- What is the parcel weight, and what does it cost to ship one unit to your main market and back?
- What is the shelf life on receipt after fulfilment minimums, not on the box?
- What percentage write-off are you budgeting for returns, breakage and expiry combined?
- Which regulatory regime does the product sit in, and who is the food business operator on the label?
- How often does a real customer finish this product and rebuy?
- What is MOQ times unit cost times number of SKUs, and how many months of runway is that?
- Have you scored at least two candidate categories against all six variables, rather than falling in love with one?
If you cannot answer those, you do not have a product idea yet — you have a product preference.
FAQ
Which food category is the most profitable to sell online?
There is no category answer, only a fit answer. Bagged herbal and fruit tea scores well on the variables that matter most online — light weight, high value density, long shelf life, low breakage and genuine repeat consumption — but it is also competitive. Dried fruit has broader natural demand and worse freight economics. Score both against your own shipping costs, capital and market before choosing.
Are high-margin products the same as profitable products?
No. Gross margin per unit is one line in a stack that also contains freight, storage, returns, breakage, expiry write-offs, channel fees and acquisition cost. A high-margin product with a heavy parcel, a short shelf life and a low repeat rate can be less profitable than a modest-margin consumable that ships cheaply and rebuys monthly.
Can I make claims about what a herbal tea does if my supplier says it is traditional?
Not on that basis. In the EU, health claims must be authorised under Regulation (EC) No 1924/2006 in the authorised wording, and supplier or folk tradition is not an authorisation. Describe the product sensorially and factually — botanical, origin, aroma, infusion colour, taste, format — and keep physiological outcomes out of your copy entirely.
How small can a first production run realistically be?
That depends on the manufacturer's line, not on your ambition. At Arovela the tea bag machine runs roughly 1,000 bags per hour, which is what makes small runs viable for us: 5,000–20,000 bags per SKU with a 2–4 week lead time, and 50,000–100,000 bags in 4–6 weeks. Dried fruit runs are agreed per fruit and format. A supplier that cannot quote a small first run is not a good partner for a first launch.
Should I start with wholesale or with my own brand?
If you have not proven sell-through, buying bulk and testing is usually cheaper than committing to branded stock nobody else can sell. If you already have a channel and want to build an asset, private label is the point. The full comparison — margin, capital, lead time, compliance — is in our wholesale versus private label guide.
Where we fit
We manufacture what we have been describing: private label tea bags in single-chamber, string-and-tag or enveloped format, and dried fruit processed at our facility in Sındırgı, Balıkesir using geothermal drying, in slices, dice, granulate, powder or purée, packed into retail formats including doypacks. Blend and recipe development, packing-only, full-service with raw material from our own range, and outer box and label printing are all in scope; quality documentation runs under our ISO 22000, ISO 9001 and ISO 27001 systems, and EU buyers can hold stock at our Solingen, Germany warehouse.
What we cannot do is tell you which category will work for you. But we can price two or three of them properly so you can score them against the variables above with real figures instead of assumptions. Request a quote with the products, formats, volumes and destination country you are considering, and say if you want a small first-run option quoted alongside a larger tier — the comparison is usually more informative than either number alone.