Define your research scope
Product research without a scope becomes noise. Start by defining the objective: a client brief, a seasonal trend report, or a team project. A clear scope focuses your product discovery and makes the final selection easier to justify and share with stakeholders.
Spot unique products with intention
To find unique products, look past generic rankings and seek curated discovery platforms that personalize to your needs. Focus on products that solve a real problem in an original way — these are the essential products that make your selection stand out and impress clients or teammates.
Identify real product trends
Product trends move fast, but real trends show sustained engagement — repeat saves, returning visitors, organic word-of-mouth — not inflated launch hype. The best products of 2026 are defined by genuine enthusiasm. Track what people actually keep and use, and your team's selection will hold up over time.
Evaluate against clear criteria
Product research only pays off if each candidate is evaluated consistently. Define shared criteria — relevance, quality, uniqueness, fit — and score every product against them. This turns subjective impressions into a defensible shortlist your team can discuss and align on.
Organize into a shareable selection
A scattered list of links isn't a selection. Group your shortlisted essential products into themed collections, tag them by use case, and add a note on why each earned its place. Tools like Merveilleux automate this — every save builds a structured, ready-to-share selection.
Share and align with your team
The final step of product discovery is communication. Share your curated selection with your team or client, explain the reasoning behind each pick, and invite feedback. A clear, organized selection turns scattered product research into a confident, collaborative decision everyone can stand behind.
Why "best" is a moving target (and why that's fine)
A word of honesty before the process: the "best products" list your team agrees on in March will not be the list in September. Products get updated, prices drift, and a tools that was perfect for last quarter's project can be wrong for this quarter's. Early in my team's attempt to build a shared selection, we treated the shortlist as a monument — and then argued about why nobody consulted it. The fix was to stop treating it as finished.
The mindset that works: your selection is a living draft with a revision date. Ours gets a fifteen-minute review at the start of each month, where the only question is "what changed?" — a price move, a new competitor, someone actually using one of the picks and reporting back. That small ritual keeps the list honest without turning maintenance into a job. A best-products list that's three months stale isn't a resource; it's a relic.
The meeting that replaced forty chat threads
Our original "process" was people dropping links into a chat channel, which sounds like collaboration and behaves like noise. A link lands, gets three emoji, and dies there — nobody can later answer the only questions that matter: who found it, why, and what happened next. We fixed it with the least glamorous change imaginable: one recurring fifteen-minute meeting where each person brings at most two finds and answers three questions about each — what problem it solves, what it costs, and what made them trust it.
Two rules made the meeting work. First, two finds maximum, which forces people to pre-filter instead of outsourcing the filtering to the room. Second, no sharing links in chat between meetings — if it's worth the team's time, it's worth fifteen days of patience. The surprising result wasn't the meeting itself; it was what happened to the chat. Almost silent. And when someone does post there now, everyone reads it, because it stopped being a landfill.
Scoring without pretending to be scientific
Every guide tells you to "evaluate against clear criteria," and every team that tries it builds a scoring spreadsheet so elaborate that by week three nobody fills it in. We did exactly this. Our criteria grid had nine dimensions and weighted averages, and it produced numbers with three false decimal places of confidence. It was astrology with formulas.
The version we kept is almost embarrassingly simple: three questions, answered in one sentence each. Does it solve a problem we actually have? Can we name someone who'd use it this month? Would we still choose it if it cost 30% more? Any "no" in the three knocks the product out unless someone argues otherwise in one sentence. That's the whole rubric. The point of scoring in a small team isn't rigor for its own sake — it's making disagreements specific enough to resolve in two minutes instead of two weeks.
What we do when we disagree
Disagreement used to stall us completely, mostly because we argued about taste as if it were fact. What fixed it was splitting the argument in two: is the product good, and is it good for us? The first is answerable with evidence — reviews, track record, comparisons. The second is answerable only by our own context — our budget, our projects, the person who'd live with the product daily. Most of our arguments turned out to be the second kind wearing the first kind's clothes.
When we still disagree after splitting it, we use the cheapest possible tiebreaker: a two-week trial with one named owner. The owner uses it, everyone else ignores it, and at the end the owner reports one sentence. It's remarkable how many debates dissolve when someone has to actually live with the product instead of speculating about it. Ownership converts opinions into evidence, and evidence ends arguments.
Where a tool earns its keep (and where it just adds another tab)
You'll notice this guide is light on tool recommendations, and that's deliberate — most teams' discovery problem isn't a missing tool, it's an unowned process. But once your process exists, the right tool removes its most boring parts. For us that meant: a shared library instead of scattered bookmarks, automatic capture of who found what and why, and a feed that pre-filters by the categories we actually work in. We didn't need software to think for us; we needed it to remember for us.
The test I'd give any discovery tool: if it vanished tomorrow, would your team lose information, or just convenience? Losing convenience is fine — that's a nice-to-have. Losing information means your knowledge only lived in the tool, which is a risk disguised as efficiency. Whatever you use, make sure the reasoning behind each pick — the three sentences from the scoring step — lives somewhere that survives the tool.
Where to start tomorrow morning
If you take nothing else from this guide, take the sequencing. Don't start by choosing a platform or building a taxonomy — start with the fifteen-minute meeting and the three-question score. Both need zero setup, and both will teach you what your team actually needs from any tool or system you add later. Process first, tool second, and only as much of either as your team will genuinely keep up.
The teams I've seen get consistently good at spotting the best products aren't the ones with the most sophisticated setup. They're the ones whose system survived contact with a busy month. Build for the busy month: small meeting, short rubric, named owners, a monthly review. It's less impressive on a whiteboard and dramatically better in practice.