Lists of "hot products" tend to go stale fast. The durable skill is scoring your own idea against the filters that decide whether it can be made and sold at a profit. Here's that scorecard — and what's actually practical to make first.
The short answer: there's no single best product idea, and a listicle of trending products is the wrong thing to chase. The best idea is the one that is best for you — it solves a real, specific problem for a group you can reach, it's different enough to choose over what exists, you can make it at a cost that leaves a margin, and you can find customers affordably. Below is a 5-filter scorecard with a stop rule for each filter, so you can score your own idea in an afternoon instead of guessing.
Run them cheapest-to-kill first. The filters are ordered on purpose — the early ones (problem, demand, differentiation, reach) cost only your time and free research, while the last one, unit economics, needs a real manufacturing quote. So kill a weak idea with the free filters before you spend anyone's time pricing a tool. Each filter here has an under-an-hour kill test you can run from your desk.
Not from brainstorming in the abstract. The ideas most likely to work come from places where you already have knowledge or friction. Three reliable veins: a frustration you live — something in a hobby or routine that annoys you and probably annoys others; a product that disappoints — the one- and two-star reviews of existing products repeat the same complaints, each one a gap; and a field you know — if you work in an area, you see needs outsiders miss. The common thread is that you understand the buyer, which makes every later step easier. For a jump-start, our Idea Spark tool turns an interest or a problem into a few concrete directions.
Score a candidate idea on each filter. The stop rules matter as much as the passes — a single hard fail is a reason to fix or drop, not to push ahead.
Does it solve a specific problem, or is it just "neat"? Neat isn't a business.
Pass: you can state the problem and who has it in one sentence. Stop if: the best case you can make is "people might like it". Five-minute kill test: write the one-sentence problem statement and who has it — if you can't, stop here.
Size it with free signals before spending: search interest for the problem and existing products; the volume and intensity of one- and two-star competitor reviews; the size of the communities where the buyer gathers (subreddits, forums, groups, hashtags); and whether similar products sell and restock on marketplaces.
A rough size check, in words: reachable audience × a plausible conversion rate should point to enough buyers to matter — if the whole findable audience is a few hundred people, the math rarely works. Also count the competitors whose reviews show the same unresolved complaint: several is a strong demand signal. One-hour kill test: spend an hour searching the problem, skimming competitor reviews, and finding the communities — if you come up empty-handed, that's your answer.
Pass: multiple independent signals (search, reviews, community size, marketplace restocks) agree a real, findable group exists. Stop (hard fail): you can't find anyone searching, complaining, or gathering around the problem — that's a drop, not a fixable gap.
Map what already exists, then name your edge. A differentiator is something a buyer can feel: cheaper, better made, or a missing feature added — not just "mine is nicer". A better version of a proven product is often a stronger bet than something brand-new.
Pass: you can finish "buyers would choose this over X because ___". Stop if: the honest answer is "it's basically the same". One-hour kill test: read the top competitors' listings and reviews, then finish that sentence out loud.
A product can pass the first three filters and still fail if the only way to find customers is expensive advertising against big brands. Ideas tied to a specific community — a sport, a trade, a hobby — tend to win here, because the buyers already gather in known places you can reach cheaply.
Pass: you can picture exactly where the first hundred customers come from. Stop if: the only answer is "run ads and hope". One-hour kill test: name your first 100 buyers and the exact places (communities, shops, events) you'd reach them — if you can't, reach is your weak filter.
The margin test, in words: selling price − landed cost − channel fees − tooling amortized per unit = per-unit profit. Landed cost is the part plus packaging plus freight and duty to your door. Channel fees are the cut a marketplace or retailer takes. Tooling amortized per unit is the one-time mold cost spread over the units you expect to make over the product's life — but you still have to fund the first run, so a small opening order makes each early unit carry a big slice of tooling. That's why a low minimum order quantity (MOQ) and a simple single-part design matter for a first product; a bridge or lower-cavity tool can cut the entry cost while you prove demand. Two costs people forget: returns/damage (budget for them) and customer-acquisition cost (what filter 4, reach, is really measuring) — both come out of the same margin. A useful target is thinking in ratios: aim for a healthy share of the selling price left as profit after every subtraction, not a few cents.
Worked in words: if a part sells for a price P, and landed cost is roughly a third of P, the channel takes a cut, and tooling-per-unit is small once spread over a real lifetime volume, you can still be left with a meaningful slice of P as profit — that's a pass. If landed cost alone is most of P, no amount of volume saves it — that's a hard fail.
Pass: a sensible price still leaves profit after all four subtractions at an MOQ you can afford. Stop if: the margin only works at a volume you can't fund, or the tool cost dwarfs a realistic first run (you amortize the tool over the product's life, but you still have to pay for that first run up front). Needs-a-quote test: this is the one filter you can't finish at your desk — a rough quote from a maker is what confirms it, which is why you run it last.
(Protecting the idea is a parallel question, not a filter — an NDA before you share it with a maker, and a prior-art search / patent decision with an attorney if copycats are a real risk.)
The idea: a small clip that keeps earbuds (the wireless kind that fall out) from dropping during a run. Problem? Yes — runners genuinely lose earbuds; the complaint shows up across fitness forums and earbud reviews. Demand? Runners gather in specific subreddits, running clubs, and Strava-type communities, so the group is real and reachable. Differentiation? Only if it fits more earbud shapes or grips better than what exists — so that becomes the design goal. Reach? Yes — those same running communities are cheap to reach, so the free filters all pass. Unit economics? Promising — it's a small part, so tooling is modest. One nuance: if it needs a soft grip, that's a second material (an overmold), which nudges it a step up from the very easiest single-material category — still friendly, just worth pricing. It clears all five, and because it's a small part it lands in the easiest manufacturing category below.
The idea: a water bottle with a screen and an app that tracks hydration. The problem is real, but it fails the unit-economics filter for a first-timer: many parts, electronics, a battery, and a regulated charging circuit mean high tooling, high landed cost, and a big MOQ before the margin works. The fix: strip it to a version one — a well-designed plain bottle with a simple measurement marking, or a single molded cap accessory that adds the useful bit — prove demand, then add electronics later once the product and the buyer are real. Same idea, re-scoped to pass the filter it was failing.
Some product types are far friendlier to a first-time maker. Practical first products tend to be a single part or small assembly made by a well-understood process: an injection-molded part (housings, organizers, clips, tools), a cut-and-sewn soft good (bags, pouches, covers), a simple enclosure or case, packaging, or a small electronics product in a molded housing. The process is known, the tooling is a defined investment, and the design can be reviewed for manufacturability before you commit. What to avoid as a first product: exotic materials, heavily regulated categories, and large multi-part assemblies — not impossible, just costlier and riskier before you've proven anything.
Sort your own idea with this quick self-test — each answer on the right nudges your idea toward harder and costlier to make first:
Mostly left-column answers mean a friendly first product. Mostly right-column means consider a simpler version one and add complexity later, once the product is proven.
When you take a surviving idea to a maker, having these ready turns a vague chat into a real quote path: a sample, sketch, or CAD of the part; how it's used and what it must withstand; your target quantity (and realistic lifetime volume); any material or finish preference; and a signed NDA. You don't need all of it to start — but each piece you bring sharpens the cost and the MOQ — the numbers filter 5 (unit economics) turns on.
Whatever idea clears the scorecard, prove demand cheaply before you spend on tooling — talk to buyers, take pre-orders, or put up a simple landing page. Then get a real manufacturing cost by having the part reviewed, so you know the margin actually exists (see what it costs to manufacture). Only then invest in tooling. That sequence is the difference between a product business and an expensive hobby. As a manufacturing network that sources worldwide, JCH can price your part across the right process and partner and tell you the realistic tool cost and MOQ — the numbers filter 5 (unit economics) turns on.
Tell Jack the idea and who it's for — he'll tell you honestly whether it's a sensible first product to make and what it would take. Or try Idea Spark for directions. NDA first, no job too small.
Book a 30-min callTry Idea SparkThere's no single best idea, and "hot product" lists go stale fast. The best one is best for you: it solves a real problem for a reachable group, is different enough to choose over what exists, can be made at a margin, and has buyers you can reach affordably. Start from a problem you understand, then score it on the five filters.
Score it on five filters and stop if it fails one hard: (1) real problem, (2) demand you can find in real numbers, (3) differentiation vs competitors, (4) reach — you know where the first hundred buyers come from, (5) unit economics — price minus landed cost, channel fees, and tooling-per-unit still profits at an affordable MOQ. Run the free filters first; pay for a quote on 5 last. Clear all five before tooling.
Use free signals: search interest for the problem and existing products; the volume and strength of 1–2 star competitor reviews; the size of the communities where buyers gather; and whether similar products sell and restock on marketplaces. Together they tell you fast whether a real, reachable group has the problem — before you pay for tooling.
A mold is a one-time cost spread across the units you make, so a tiny first run makes each part carry a big slice of tooling and can erase the margin. A low MOQ and a simple single-part design keep the entry affordable; a bridge or lower-cavity tool can lower it further while you prove demand. A manufacturer can tell you the realistic MOQ and tool cost for your part.
A single part or small assembly made by a known process — an injection-molded part, a cut-and-sewn soft good, a simple enclosure, packaging, or a small electronics product in a molded housing. Avoid exotic materials, regulated categories, and large complex assemblies for a first product. A maker can tell you if a simpler version launches faster and cheaper.