Confluence LabsYC W26 landing page report.
← Back to YC W26
Confluence Labshttps://confluence.sh/

Your landing page’s weakest score is CTA.

Fix CTA

Score over time

6050403020Sakana AI 44 on Jun 23 (Wayback)Sakana AI 43 on Jul 12 (Wayback)Sakana AI 45 on Jul 24 (Wayback)Sakana AI 46 on Aug 13 (Wayback)Sakana AI 44 on Aug 21 (Wayback)Sakana AI 46 on Sep 9 (Wayback)Sakana AI 43 on Sep 20 (Live)Basis 55 on Apr 19 (Wayback)Basis 53 on May 19 (Wayback)Basis 51 on Jun 15 (Wayback)Basis 52 on Jul 21 (Wayback)Basis 50 on Sep 22 (Live)Lila Sciences 33 on Apr 22 (Wayback)Lila Sciences 33 on May 4 (Wayback)Lila Sciences 32 on Jun 4 (Wayback)Lila Sciences 38 on Jun 11 (Wayback)Lila Sciences 37 on Jul 23 (Wayback)Lila Sciences 40 on Sep 12 (Wayback)Lila Sciences 41 on Sep 20 (Live)Periodic Labs 53 on May 28 (Wayback)Periodic Labs 50 on Jun 10 (Wayback)Periodic Labs 51 on Jun 30 (Wayback)Periodic Labs 51 on Jul 25 (Wayback)Periodic Labs 51 on Aug 27 (Wayback)Periodic Labs 50 on Sep 13 (Wayback)Periodic Labs 53 on Sep 20 (Live)Radical AI 46 on Jun 10 (Wayback)Radical AI 48 on Jun 25 (Wayback)Radical AI 46 on Aug 8 (Wayback)Radical AI 49 on Sep 20 (Live)Confluence Labs 52 on May 18 (Wayback)Confluence Labs 50 on Jun 9 (Wayback)Confluence Labs 52 on Sep 21 (Wayback)Confluence Labs 53 on Sep 22 (Live)May 18Jun 9Sep 21Sep 22
Past reportsShow past reportsHide past reports
  1. Sep 22, 202653LiveOverall +1 · Trust +6 · Information hierarchy −4Viewing
  2. Sep 21, 202652Overall +2 · Information hierarchy +6 · Trust +4View →
  3. Jun 9, 202650Overall −2 · Information hierarchy −6 · Trust −4View →
  4. May 18, 202652First tracked report — strongest Information hierarchy 58, weakest CTA 44View →

CTA

42Poor

Copy, size, verb, cost-feel — does the primary CTA earn the click?

founders@confluence.sh
Confluence Labs
Lila Sciences
Introducing Basis ↗
Basis
Periodic Labs

Problems

  • no button or input at all
  • mailto is a high-cost click
  • inline link has zero visual weight
  • no action verb — reads as a label

Mobile web

52Fair

The phone experience as a first-class citizen.

Mobile web annotations on the mobile capture1
  1. Buried writing link — The only tappable element wraps badly, stranding "2" on its own line, and is a thin text link, not a thumb target.

Information hierarchy

54Fair

Does the design help or inhibit what the page is trying to say?

Information hierarchy annotations on the desktop capture123
  1. Four-paragraph text wall — Four same-weight body paragraphs stack with no visual anchor, forcing the visitor to read linearly instead of scanning.
  2. Muted tagline text — "AI models that learn from experience" is set in low-contrast grey-blue, subordinating the value prop to the body copy below it.
  3. Footer inside fold — Footer logo and credits creep into the first viewport, signaling the page is over before any action is asked for.

Value props

58Fair

Title, subtitle, hero — do they sell the product's true value?

Confluence Labs’s headline undersellsConfluence LabsAI models that learn from experience.

Working

  • Body copy names real mission clearly
  • Big emotional stakes: cure disease, fusion
  • YC and Paul Graham backing as proof

Could improve

  • Subhead is generic AI-category filler
  • SOTA on ARC-AGI-2 buried as a link
  • No numbers or outcomes above headline fold

Trust

58Fair

Proof that the product is real, competent, and safe to try.

Trust read
Confluence LabsBacked by Y Combinator and Paul Graham
Sakana AIimplicit context Sakana AI may benefit from pre-existing familiarity among its target audience.
BasisNo visible trust signal identified.
Lila SciencesNo visible trust signal identified.
Periodic LabsNo visible trust signal identified.
Radical AINo visible trust signal identified.

Competitors

inferred from product, audience, and value prop