Product Launch Statistics 2026: 30+ Data Points From 57,000 Launches
October 17, 2026 · Nicheloom
On a typical day this autumn, about 1,200 new products launched across Product Hunt, GitHub, Hacker News and Y Combinator. Roughly 3 in 10 were AI-native, about 6 in 10 were commercial products, and Tuesday was the busiest launch day while Sunday was the quietest.
Those numbers come from Nicheloom's own daily tracking: 57,422 launches in total, and 34,230 launches in the four complete weeks from September 14 to October 11, 2026. Every table below says how many days or launches it rests on, so you can judge how far to trust it.
The numbers at a glance
| Question | Answer | Based on |
|---|---|---|
| Launches per day (median) | 1,204 | 28 days, 34,230 launches |
| Busiest weekday | Tuesday, 1,512 per day | 4 weeks |
| Quietest weekday | Sunday, 889 per day | 4 weeks |
| AI-native share of launches | 30.4% | 33,825 classified launches |
| Launches with any AI angle (native or feature) | 39.1% | 33,825 classified launches |
| Commercial products vs hobby or open source | 61.8% vs 38.2% | 33,890 classified launches |
| Launches that look vibe-codeable | 62.4% | 34,226 assessed launches |
| Largest niche in the window | AI API gateways and proxies, 352 launches | 28 days |
| Active niches we track | 820 | all launches |
| Median upvotes on a Hacker News launch | 8 points | 568 launches |
| Median stars on a GitHub launch | 19 | 9,974 launches |
| SEC Form D filings per month | about 550 to 580 | July to September 2026 |
How many products launch per day?
About 1,204 per day across the four launch sources we track, with a range from 818 on the quietest day to 1,661 on the busiest. Almost all of that volume comes from two places.
| Source | Launches in 28 days | Share of all launches | Median per day |
|---|---|---|---|
| Product Hunt | 23,632 | 69.0% | 816 |
| GitHub | 9,974 | 29.1% | 328 |
| Hacker News (Show HN) | 568 | 1.7% | 20 |
| Y Combinator | 56 | 0.2% | 1 |
| All four | 34,230 | 100% | 1,204 |
Product Hunt alone is seven in ten launches, and most of those submissions are never featured. Hacker News and Y Combinator are small by comparison. For how the two compare, see Hacker News vs Product Hunt: first launch data.
Which day of the week has the most launches?
Tuesday. Averaged over four weeks, Tuesday had about 1,512 launches, around 1.7 times Sunday's 889. Product Hunt is even more lopsided: 1,118 on a Tuesday against 550 on a Sunday, about 2 times.
| Day | All sources, launches per day | Product Hunt only, launches per day |
|---|---|---|
| Monday | 1,188 | 731 |
| Tuesday | 1,512 | 1,118 |
| Wednesday | 1,273 | 928 |
| Thursday | 1,282 | 888 |
| Friday | 1,351 | 954 |
| Saturday | 1,063 | 740 |
| Sunday | 889 | 550 |
Four weeks is a short sample, so treat the exact figures as directional. The shape (a midweek peak and a Sunday low) is what to take from it. If you are choosing a launch day, how many upvotes you need on each day of the week matters as much as how many other launches you compete with.
What share of new launches are AI?
Across all four sources, 30.4% of launches were AI-native (AI is the core of the product), another 8.7% added an AI feature, and 60.8% had no AI angle. Put together, 39.1% of launches involve AI in some way.
It differs a lot by where the launch appeared:
| Source | AI-native | AI feature | Not AI | Launches classified |
|---|---|---|---|---|
| Y Combinator | 83.9% | 5.4% | 10.7% | 56 |
| Hacker News | 38.3% | 6.0% | 55.7% | 567 |
| GitHub | 34.5% | 8.4% | 57.1% | 9,574 |
| Product Hunt | 28.5% | 8.9% | 62.6% | 23,628 |
Y Combinator's figure rests on only 56 launches, so read it as "most", not as a precise percentage.
Is the AI share rising?
Among Hacker News and Y Combinator launches, the only two sources we have tracked back to late October 2025, the AI-native share was steady from March to August and then jumped:
| Month | AI-native share | Launches |
|---|---|---|
| March to August 2026 | 27% to 32% each month | 727 to 925 per month |
| September 2026 | 36.9% | 680 |
| October 2026 (to the 16th) | 42.2% | 365 |
That is the clearest trend in our data, but it is two sources and two months, so it could also reflect who posted during a busy stretch. Across all four sources, the weekly AI-native share drifted from 32.6% to 27.3% over the four weeks. The two views point in different directions, so we will keep tracking it monthly before calling a trend.
How many launches are real businesses?
61.8% of launches were commercial products, and 38.2% were hobby or open-source projects. The split depends almost entirely on the source:
| Source | Commercial | Hobby or open source |
|---|---|---|
| Y Combinator | 98.2% | 1.8% |
| Product Hunt | 83.2% | 16.8% |
| Hacker News | 24.8% | 75.2% |
| GitHub | 11.5% | 88.5% |
So a Show HN post is usually a project someone built for fun, while a Product Hunt launch is usually trying to be a business. If you want to study what people are trying to sell, Product Hunt and Y Combinator are the cleaner data. Hacker News and GitHub show what builders make for themselves.
Who are new launches built for?
| Audience | Share of launches |
|---|---|
| Consumers (B2C) | 29.7% |
| Prosumers | 27.8% |
| Developers | 25.3% |
| Businesses (B2B) | 17.2% |
Based on 33,646 classified launches. B2B is the smallest group at 17.2%. For where specific industries are thinly served, see our market gaps data.
What do new launches actually do?
The most common jobs, by share of 33,218 classified launches:
| What it does | Share |
|---|---|
| Dev tools | 21.9% |
| Workflow automation | 19.6% |
| Content generation | 14.0% |
| Analytics and BI | 7.5% |
| Agents and copilots | 7.4% |
| Vertical SaaS | 6.0% |
| Search and retrieval | 4.3% |
And by industry, 41.8% of launches are horizontal (not tied to one industry) and 20.9% are in media and entertainment. After that, the next largest are marketing (5.8%), fintech (5.6%), education (4.9%) and retail and commerce (4.4%). Healthcare is 3.1% and HR and recruiting 2.0%. Each specific industry holds a small share of launches.
How many launches could be vibe coded?
Our classifier judged 62.4% of launches as something a solo builder could reproduce with an AI coding tool, 19.5% as partly, and 18.1% as no. By source, 65.2% of Product Hunt launches, 56.4% of GitHub launches and 52.5% of Hacker News launches were judged reproducible (Y Combinator was 7.1%, but on only 56 launches). For the method and what separates the three groups, read how we scored 42,000 launches on vibe-codeability, and for the effect on attention, vibe-codeable launches get half the upvotes.
Which niches are biggest, growing and shrinking?
We group launches into 820 niches, and 819 of them had at least one launch in the window. The ten largest by launches in the four weeks:
| Niche | Launches |
|---|---|
| AI API gateways and proxies | 352 |
| Developer utility tools | 329 |
| AI video generation tools | 278 |
| Browser-based utility tools | 263 |
| AI agent development platforms | 217 |
| Model context protocol tools | 191 |
| Specialized calculation tools | 183 |
| Generative engine optimization software | 171 |
| Language learning apps | 170 |
| macOS system utilities | 168 |
What grew fastest?
Comparing the first two weeks of the window with the last two, among niches with at least 15 launches in the first half and 20 in the second:
| Niche | First 2 weeks | Last 2 weeks | Change |
|---|---|---|---|
| Game console homebrew tools | 17 | 60 | 3.5 times |
| Video game source ports | 40 | 125 | 3.1 times |
| Specialized algorithms and computing libraries | 29 | 75 | 2.6 times |
| Computer-aided design software | 21 | 53 | 2.5 times |
What fell?
Among niches with at least 25 launches in the first two weeks: machine learning inference engines went from 39 to 11 (0.28 times), developer documentation platforms from 31 to 9, and speech and text conversion software from 25 to 9.
Two weeks against two weeks is a noisy comparison, and a cluster of related open-source launches can swing a small niche. Use these as places to look, not as forecasts. The fastest-growing SaaS niches by launch volume post covers a longer window.
How many upvotes does a launch get?
Upvotes mean something different on every source, so there is no single answer.
| Source | Measure | Median | 90th percentile | Share at 100 or more |
|---|---|---|---|---|
| Hacker News | points | 8 | 93 | 9.2% |
| GitHub | stars | 19 | 108 | 11.3% |
Both are based on launches in the 28-day window (568 and 9,974 launches). On Product Hunt, most submissions are never featured and get about 1 upvote, so the useful numbers are for featured launches: a median of 80 upvotes, about 170 to reach the top 5 and about 105 for the top 10, as of October 14. The full breakdown is in how many upvotes you need on Product Hunt.
How much private funding is being filed?
We also track SEC Form D filings, the notice companies file when they raise money privately. Our tracker has covered them since June 11, 2026.
| Month | Filings | Median amount on the filing |
|---|---|---|
| July 2026 | 575 | about $2.40 million |
| August 2026 | 546 | about $2.49 million |
| September 2026 | 578 | about $2.50 million |
Form D filings include funds and other private placements, not only software startups, and the amount is what the filing reports, so use it as a gauge of private fundraising activity rather than of startup funding specifically. October is partial (357 filings by the 16th) and June is partial (tracking began on the 11th).
How we counted
- Source of the data: Nicheloom collects launches daily from Product Hunt, GitHub, Hacker News (Show HN) and Y Combinator, plus SEC Form D filings for funding.
- Main window: the four complete weeks from September 14 to October 11, 2026 (28 days, 34,230 launches). We chose it because all four launch sources were being collected in full for those weeks. Product Hunt became a full feed in early September (before that we stored only its top launches each day), and GitHub began on September 9.
- Longer history: only Hacker News and Y Combinator go back to late October 2025, so the monthly AI trend uses those two.
- Classification: each launch is labelled by an AI model that reads its name, tagline and description (AI stance, commercial or hobby, audience, function, industry and vibe-codeability). These labels are good for large-sample patterns but can be wrong on any single launch, and counts exclude launches not yet classified, which is why the base differs between tables.
- What counts as a launch: a new listing on one of the four sources. The same product posted on two sources counts twice.
- Upvotes: points on Hacker News, stars on GitHub, votes on Product Hunt. They are not comparable across sources.
- Excluded: launches removed after a takedown request are not counted.
We will refresh this page monthly.
What to do with these numbers
- Check the base before you quote a share. "30% of launches are AI" is true across all four sources, but 28.5% on Product Hunt and 83.9% (of a small group) at Y Combinator.
- Pick your launch venue for your goal. Product Hunt has the volume and the most commercial products. Hacker News is mostly projects, with a small number getting large attention.
- Look at niches, not categories. The biggest niches are crowded with dozens of launches a week, while a niche with a handful of launches and rising interest is a better place to start. Browse them on our niches page, or see what launched in the last day on the Today page.
- Use the gaps. Where launches are thin (healthcare is 3.1% and HR and recruiting 2.0%), our market gap matrix shows which industry and function pairs have the fewest products.
Data as of October 17, 2026, from Nicheloom's own tracking.
Frequently asked questions
How many products launch per day? About 1,200 across Product Hunt, GitHub, Hacker News and Y Combinator, based on a median of 1,204 launches per day over the four weeks to October 11, 2026. The range was 818 to 1,661. Product Hunt contributes about 816 a day and GitHub about 328.
What percentage of new product launches are AI? 30.4% are AI-native and another 8.7% add an AI feature, so 39.1% involve AI. The share varies by source, from 28.5% on Product Hunt to 38.3% on Hacker News. Among Hacker News and Y Combinator launches, the AI-native share rose from 27% to 32% a month through August to 42.2% in October 2026.
What is the busiest day for product launches? Tuesday, with about 1,512 launches a day averaged over four weeks. Sunday is the quietest at about 889. On Product Hunt alone, Tuesday has about 1,118 launches and Sunday about 550.
How many launches are real businesses? 61.8% are commercial products and 38.2% are hobby or open-source projects. On Product Hunt 83.2% are commercial, at Y Combinator 98.2%, on Hacker News 24.8% and on GitHub 11.5%.
What are the biggest niches for new products right now? By launches in the four weeks to October 11, 2026: AI API gateways and proxies (352), developer utility tools (329), AI video generation tools (278), browser-based utility tools (263) and AI agent development platforms (217).
How many upvotes is a good launch on Hacker News or GitHub? On Hacker News the median launch gets 8 points and about 1 in 11 reaches 100. On GitHub the median is 19 stars and about 1 in 9 reaches 100. For Product Hunt, see the dedicated benchmarks post.