What to sell online: the tests a product needs to pass

Nobody can tell you what to sell online

Every "top 50 trending products" list is out of date by the time it ranks. The viral fidget spinner of 2017 is the LED strip of 2020 is the "viral TikTok" of last month. By the time a trend shows up in a listicle, the margins have been competed away.

Here's what actually works for picking products to sell online, based on the patterns we see from Shoprocket stores that make it past their first year.

The three tests a product needs to pass

1. You have some unfair edge finding or making it

If you're picking products from the same AliExpress catalogue everyone else is looking at, you have no edge. Your competition is identical, your margin is whatever the paid-ad auction will let you keep.

Real edges look like:

• A supplier relationship nobody else has (local manufacturer, niche distributor, family connection)
• A craft skill that's hard to replicate (custom engraving, hand-finishing, specific material expertise)
• Geographic arbitrage (an item that's cheap where you are, rare where you're selling)
• A distribution channel nobody else has (your own audience, your own community, your own content)

If your answer to "why you?" is "because I saw it was trending on Twitter", the answer is nothing.

2. The margin survives contact with reality

A lot of ecommerce advice glosses over the true cost of selling online. At minimum, you're paying:

• Payment gateway: 2.9% + 30¢ per transaction (Stripe, PayPal)
• Platform fees: 0% on Shoprocket; 2% on Snipcart; 10% + 50¢ on Gumroad; variable on Shopify; 15% on Amazon FBA
• Shipping: highly variable; often eats 5-15% of revenue unless baked into price
• Returns and refunds: typically 5-10% of sales volume for physical goods
• Customer acquisition: $5-50+ per customer on paid ads, less if you have organic traffic
• Your time: packing, support, admin (don't forget this)

A product that sells for $30 and costs $10 to source looks like a 66% margin on paper. After the list above, the real margin is often 10-15%. If it's thinner than that, volume has to be huge for the business to work.

3. There's a real audience, not just a search trend

Google Trends can tell you a topic is rising. It can't tell you whether the people searching have money and intent to buy. A trending keyword can be journalists writing about a product, curious teenagers, or bots.

Better signals:

• An active niche community (Reddit, Discord, hobby forums) discussing the problem the product solves
• Existing competitors actually running paid ads on the term (ads cost money; their existence means somebody's converting)
• Repeat-buy behaviour, not one-and-done purchases

Product categories that hold up over time

Rather than a trending list that dates itself in six months, here are categories that consistently make sense for independent ecommerce:

Consumables

Anything people buy repeatedly. Coffee, tea, supplements, skincare, pet food, cleaning products. Subscription-friendly, predictable revenue, strong LTV.

Craft and handmade

Products where you are the edge. Leather goods, ceramics, prints, candles, jewellery, custom clothing. Margins are usually healthy because buyers aren't comparison-shopping on identical SKUs.

Digital

Ebooks, templates, courses, software, music, Lightroom presets, Figma kits. Near-zero marginal cost, instant delivery, global market. The hard part is the audience, not the product.

Niche physical goods

Categories that are too small for Amazon to target seriously: specialty hobby supplies, niche sports gear, music equipment accessories, vintage items, collectibles. If the total addressable market is $10M/year globally, you can own a meaningful slice.

Bundles and collections

Take individual products, package them into a themed bundle (a "starter kit" for a hobby, a curated box, a pairing set). Average order value goes up, perceived value goes up, margin usually holds.

What we see working on Shoprocket stores

From stores we watch grow on Shoprocket specifically:

• Creators selling digital products (templates, courses, one-off downloads) alongside blog or social content they already produce
• Hobby suppliers (3D printing filament, watch straps, miniature paints) with strong niche communities
• Specialty food and drink (single-origin coffee, craft soda, specialty sauces)
• Custom / made-to-order products (engraved items, bespoke prints)
• Recurring consumables sold via subscription

The common thread isn't the product. It's that the seller has a distinct reason their version of that product should exist in the market, and an audience (or channel) that knows it.

Categories that don't tend to work

• Generic dropshippable products: oversaturated, no edge, thin margin, high ad cost
• Me-too fashion: unless you're a brand, you're competing with Amazon + Shein on price
• Low-ticket one-off items: the fulfilment economics don't work under ~$20 order value unless you batch
• Trending gadgets with a 6-month shelf life: by the time you've sourced inventory, the trend is gone

How to actually start

1. Pick a category where you have a real edge (relationship, skill, channel, audience)
2. Source or create 3-5 products in that category
3. Launch with a minimal storefront: a single-page site or an embed on a site you already have
4. Take real orders from real people before investing in inventory scale
5. Let feedback shape the next 10 products

For the storefront part, Shoprocket is a one-script embed on any site, zero platform transaction fees, built for the "don't rebuild your site, just add commerce" path:

<!-- EXAMPLE ONLY. Grab your real snippet from Sales channels → Embeds in your dashboard -->
<script src="https://cdn.shoprocket.io/loader.js" data-pk="pk_yourkey"></script>
<div data-shoprocket="catalog" data-embed-id="emb_xxx"></div>

14-day trial, no card required. Start free. Pricing starts at $29.

TL;DR

• Trend lists are a trap: by the time you find them, the margin is gone
• Every product needs: your unfair edge + real margins after all costs + an actual audience
• Categories that age well: consumables, handmade, digital, niche physical, bundles
• Start small: 3-5 products, minimal storefront, real feedback loop, scale what works