The verified organiclaunch directory

A launch appears here only after its reading came back organic. That is the whole entry criterion, and it is deliberately narrow: 10 of the 30 launches RADAR tracks clear it. The other 20 are read, published, and counted in the distribution above, but they are never named on this page. A listing is therefore a positive result, not a directory entry every launch gets by existing.

Each listing also carries a letter, and the letter is about evidence rather than performance. An A means the organic read sits on a full forensic trace of how the reach arrived. A B means the same organic read was reconstructed from the published ratio and the engagement shape without that trace. Per our own trace records, 7 of the 10 launches named here carry an A and 3 carry a B. A B is not a weaker launch. It is a weaker evidence file, and saying so is more useful than rounding it up.

Nothing on the page is a claim about a product, a company, or a founder. The products are real and the launches happened; what RADAR reads is how the reach was built. The full method, the thresholds, and the confidence rules are written out on the RADAR methodology page, and the ranked version of this same safe set is the RADAR leaderboard.

Views per like is the load-bearing signal, but it is not the verdict. Across the named set the ratio runs from 98 to 2,709 with a median of 396, and the top of that range sits well above the roughly 500 views per like RADAR treats as the organic ceiling. One launch on this page clears the ceiling by a wide margin and still reads organic, and that is worth explaining rather than hiding.

The reason is that a full reading weighs four more things. Whether reach arrived on a curve that engagement followed, or in a block engagement never caught up with. Whether the written layers grew in step with the views. Whether the accounts replying and quoting read as real and spread out rather than as one cluster. And whether the quote and repost pattern looks like word of mouth or like a known amplifier ring. A very large, established account distributing to an audience far beyond its own followers produces a high ratio for an ordinary reason: most of those people scroll past without tapping like.

So the ratio is where a reading starts and not where it ends. That cuts both ways. A launch can sit comfortably inside the ceiling and still fail the coupling or the reply-network check, in which case it is not named here either. The ratio itself is benchmarked in our views-to-likes benchmark, and the wider picture of how launch videos perform sits in the state of launch videos.

The grid below has two controls: a grade filter and a sort. The filter narrows the list to the A reads (7 launches) or the B reads (3 launches), which is the fastest way to see only the entries backed by a full trace. The sort reorders whatever the filter left. Neither control ever adds a launch that was not already in the named set.

Sort by reach when you want scale. The named set runs from 62,447 views at the small end to 7,081,850 at the large end, a median of 2,527,487, so this sort spans two orders of magnitude and the top of the list is dominated by a handful of very large launches. Sort by newest when you want recency: the named launches run from January 2026 to June 2026, inside a tracked corpus that starts in October 2025.

Sort by engagement when you want the cleanest reads rather than the biggest ones. It orders by views per like ascending, so the launches whose likes kept closest pace with their views come first, starting at 98 views per like. That ordering usually disagrees with the reach sort, which is the useful part: the launch with the most views is rarely the launch with the tightest engagement coupling. For a curated read of the strongest launch videos rather than a sortable list, see the best product launch videos.

A listing says one thing: this launch's reach read as organically built. It is not a quality rating, not a ranking of the products, and not a statement that a launch not listed here did anything wrong. Absence from the list has three ordinary explanations before it has an interesting one, and the most common is simply that the launch was never in the corpus.

The size of the set is the honest constraint. Per our own corpus, 10 named launches drawn from 30 tracked is enough to show a method works and nowhere near enough to describe an industry. The selection also tilts: RADAR reads launches that put up enough public reach for a views-to-likes comparison to mean something, so a launch that reached a few thousand people is not missing because it read badly, it is missing because at that size the ratio is noise.

Read a listing as a signal that one launch cleared one published, checkable read. Read the 33% and 67% figures as a distribution of this corpus and not of launches in general. If you want the corpus-wide picture rather than the named set, the launch authenticity report is the right surface, and the method behind every figure on both is on the methodology page.