Launch

We Scored 238 Launch Pages. Here's What the Top 18% Have That the Rest Don't.

We scored 238 real launch pages across 7 dimensions. 90% had no demo. Top scorers almost all show proof, screenshots and pricing. See the data.

Published 2026-11-03.

Score your launch page freeRead the blog

Nine out of ten launch pages we scanned had no demo video. The average page scored 50.8 out of 100.

Since June, founders and small business owners have pasted their URLs into the free LaunchBuddy scan. Each page is scored on seven dimensions, and the scan names its biggest gap. After cleaning the data, we were left with 238 distinct sites. Mostly they are indie products, new apps, small online stores and service businesses, and many were built in the last few months with AI coding tools.

This post covers what we found: where launch pages lose points, what the highest-scoring pages consistently have, and the short list of fixes that close most of the distance.

How we did this

  • The sample: 238 unique domains scanned at launchbuddy.com/scan between June 23 and November 3, 2026. Where a site was scanned more than once, we kept its latest scan. We removed our own sites, internal test scans, and a handful of big-company pages people scanned out of curiosity.
  • The score: 100 points across seven dimensions: product clarity (20), demo and creative (20), conversion path (15), social share readiness (15), trust and proof (10), launch channel readiness (10) and measurement (10).
  • The checklist: 234 of the pages also have a 13-item launch checklist, with each item marked found or missing. We compared the 44 pages that scored 70 or above with the 99 under-50 pages that have a checklist (102 pages scored under 50 in all; 3 of them have no checklist).
  • The limits: this is a static scan of the homepage. It can't see content that only loads through JavaScript, and it doesn't measure sign-ups or sales. A high score means a page has the signals a cold visitor looks for. It doesn't prove the page converts. The quoted percentages are for the scanned pages only. We don't claim they represent the whole web.
  • Why the numbers differ from our homepage: this post uses the cleaned set of 238 distinct sites, after removing test scans and scans of our own sites. The live counter on the homepage counts every scanned site, so it shows a slightly larger total and slightly different percentages.

Finding 1: Most pages land in the middle, and few are fully ready

ScorePages
0–2432
25–4970
50–6992
70–8440
85–1004

The median is 55. Of the 238 pages, 102 (43%) scored under 50, 44 (18%) reached 70 or more, and 4 cleared 85.

The middle band is where most of these products sit: they're built and live, but a stranger arriving from a launch post would struggle to see what the product does, why to trust it, or what to do next.

Finding 2: Clarity is the strongest dimension. Showing and proving are the weakest.

Here is the average score on each dimension as a share of its maximum:

DimensionAvg score / max% of max
Product clarity14.4 / 2072%
Social share readiness10.1 / 1567%
Launch channel readiness5.6 / 1056%
Trust and proof4.6 / 1046%
Conversion path6.6 / 1544%
Measurement3.3 / 1033%
Demo and creative6.4 / 2032%

On the whole, founders can describe what they built. Showing it working, giving a reason to believe it, and making the next step obvious are where they fall short.

The "primary gap" the scan assigns to each page tells the same story:

  • Demo and creative: 41% of pages
  • Conversion path: 33%
  • Trust and proof: 11%
  • Social share readiness: 8%
  • Product clarity: 5.5%

Three out of four pages have their biggest problem in one of two places: there's nothing to watch, or there's no clear next step.

A note on measurement: a static scan often can't confirm analytics loaded by JavaScript, so read the measurement score in the table above as "unverified" rather than missing. For that reason we leave analytics out of the checklist comparisons below.

Finding 3: Demo videos are missing almost everywhere, including on the best pages

90% of pages had no demo video or walkthrough. Every page that scored under 50 was missing one, and so were 73% of the pages that scored 70 or above.

The demo is the one gap even the strongest pages share, which also makes it the biggest opportunity. Products built quickly with AI tools are especially hard to judge from a description: what makes them interesting is the workflow, and a paragraph of text doesn't show a workflow. Even a 20–30 second screen recording of the core action gives a cold visitor something no headline can.

Finding 4: Proof, screenshots and pricing separate the top pages from the rest

This is the clearest result in the data. Here is how often each checklist item was missing in the two groups:

Checklist itemMissing: pages scoring 70+ (n=44)Missing: pages under 50 (n=99)
Social proof / testimonials5%85%
Product screenshots / creative7%84%
Pricing or offer2%62%
Clear headline (H1)5%68%
Open Graph title + description0%41%
Social share image11%53%
Primary call to action27%77%
Signup or waitlist27%65%
Blog / changelog / updates23%86%
Demo video73%100%

Nearly every top-scoring page has at least one piece of proof: a testimonial, a usage number, a logo, a named founder. Nearly every one shows the product in a screenshot, and nearly every one says what it costs, or that it's free.

Pages under 50 usually have none of those three.

None of this is hard to build. A screenshot takes a minute. A pricing line takes one sentence. For proof, a single real quote from a beta user, or an honest "used by 40 people this month", is more than most low-scoring pages have.

Finding 5: A third of pages miss the same three things together

Across all 234 checklist pages:

  • 55% have no clear primary call to action
  • 50% have no signup or waitlist form
  • 50% show no social proof
  • 52% show no product screenshots

79 pages (34%) were missing a demo, social proof and a primary CTA all at once. For a visitor arriving from a launch post, that page offers nothing to watch, nothing to trust and nothing to click. Whatever traffic the launch brings mostly leaves without a trace.

Finding 6: Single-page landing sites scored the lowest

For the 142 pages where the scan classified the site type, average scores were:

Site typePagesAvg score
Ecommerce3766.5
Blog / content2065.2
App3959.2
Multipage site2152.8
Single landing page2528.7

That's a small sample, but the gap is large. The single landing page is the default first artifact for a new product, especially one scaffolded in an afternoon, and in this sample it was the format most likely to be missing proof, pricing and a next step. Stores and content sites are built on platforms that come with product images, prices and "add to cart" by default, which probably explains part of their lead.

Finding 7: A lot of these sites were built with AI tools

We profiled the build stack of 22 recently scanned sites. 12 of the 22 were built with a coding agent or an AI app builder: Next.js and hand-written sites with agent-style code comments, plus Base44 and Lovable/Bolt-style exports. The other 10 were on WordPress, Shopify, Wix, Weebly or hand-built.

Twenty-two sites is a small sample, but it matches what we see in the scans: building the product has become fast and cheap. The launch page that presents it hasn't caught up.

What to fix first

If you have an hour before you share your link anywhere, here's the order the data suggests:

  1. Add one screenshot of the product doing its main job, above the fold. Top pages almost never skip this: it's missing on 7% of pages scoring 70+, against 84% of pages under 50.
  2. Add one piece of real proof. A quote, a usage number, or your name and face as the builder. It's missing on 5% of pages scoring 70+, against 85% of pages under 50. Don't invent it. A small true number beats an impressive fake one.
  3. Say the price, or say it's free. Pricing is missing on 2% of pages scoring 70+, against 62% of pages under 50.
  4. Pick one primary call to action and repeat it. If you're not ready to sell, use a waitlist or email capture so launch traffic becomes a list.
  5. Record a 20–30 second demo. Even most top pages don't have one yet, so having one puts you ahead of the pack.
  6. Fix your share card. Set an Open Graph title, description and image, so the link looks like something when you post it on X, LinkedIn, Slack or Product Hunt.

For the full list the scan checks, see the product launch checklist. For the same data broken down by page type, with famous pages to learn from, see the examples for SaaS landing pages, startup landing pages, waitlist pages and product demo videos, or read the landing page best practices guide.

We'll update this study as more pages are scanned. If you'd like the numbers for an article or newsletter, link to this page and use whatever you need.

Share the study charts using the provided embed code, which links back to this study and preserves the source and sample details.

Score your own page

The scan behind this study is free. Paste your URL and in about a minute you get your score across all seven dimensions, your single biggest gap, and the specific fix for it.

If the biggest gap turns out to be something you'd rather not build yourself, like a demo video or a lead-capture page, we can make that one piece for you. Start with the score and decide from there.

Scan your launch page free at launchbuddy.com/scan

Related LaunchBuddy resources

Vibe Coding Security Checklist: 17 Things to Audit Before Real UsersHow to Get Your First Users: A Distribution Playbook for Products Built With AIProduct Launch Metrics: The Day-Two Dashboard and Daily Checks for an MVPLaunch Readiness Checklist: A 100-Point Self-Assessment for Solo Builders
We Scored 238 Launch Pages. Here's What the Top 18% Have That the Rest Don't.