Lantern AIYC F26 landing page report.
← Back to YC F26

Your landing page’s weakest score is Trust.

Fix Trust

Score over time

80706050Mercor 75 on Jun 25 (Wayback)Mercor 76 on Jul 14 (Wayback)Mercor 77 on Jul 23 (Wayback)Mercor 74 on Aug 12 (Wayback)Mercor 59 on Aug 28 (Wayback)Mercor 57 on Sep 5 (Wayback)Mercor 54 on Sep 20 (Live)Juicebox (PeopleGPT) 72 on Apr 16 (Wayback)Juicebox (PeopleGPT) 74 on May 20 (Wayback)Juicebox (PeopleGPT) 75 on Jun 24 (Wayback)Juicebox (PeopleGPT) 69 on Jul 1 (Wayback)Juicebox (PeopleGPT) 75 on Jul 25 (Wayback)Juicebox (PeopleGPT) 69 on Sep 2 (Wayback)Juicebox (PeopleGPT) 77 on Sep 20 (Live)Dover 64 on Mar 6 (Wayback)Dover 65 on Apr 16 (Wayback)Dover 68 on May 17 (Wayback)Dover 68 on May 27 (Wayback)Dover 68 on Jun 18 (Wayback)Dover 63 on Jul 25 (Wayback)Dover 69 on Aug 3 (Live)Paraform 75 on May 13 (Wayback)Paraform 75 on Jun 3 (Wayback)Paraform 77 on Jul 6 (Wayback)Lantern AI 68 on Sep 20 (Live)Sep 20

Trust

48Poor

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

Lantern AI
Juicebox (PeopleGPT)
Paraform
Dover
Trust read
Lantern AIBACKED BY Y COMBINATOR · F26 badge
Mercorimplicit context Mercor may benefit from pre-existing familiarity among its target audience.
Juicebox (PeopleGPT)"58% OF THE FORBES AI 50" logo band with Cursor, Harvey, Notion, Carvana, EPAM logos
Dover"Unmatched hiring transparency" table stating "This table shows every hire and the total cost to the company"
ParaformCustomer logo wall featuring Palantir, Rippling, Decagon, Abridge, Cognition, Owner, and Scale

CTA

68Fair

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

Lantern AI
Juicebox (PeopleGPT)
Paraform
Dover

Problems

  • demo booking feels costly
  • input funnels to calendar, not product
  • no zero-commitment path visible

Information hierarchy

74Fair

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

Information hierarchy annotations on the desktop capture1234
  1. Duplicate Book a Demo — Two 'Book a Demo' CTAs in the fold — the orange nav button and the cream form button compete at different weights, splitting the scan path.
  2. Eight-item nav row — JUDGMENT through FAQ adds eight same-weight exit links directly above the hero, each one a leak before the primary action.
  3. Human/LLM toggle — A novel 'READ AS Human/LLM' control sits at the same visual weight as the primary CTA and demands processing before its purpose is clear.
  4. Form CTA mismatch — The input asks 'Who do Lantern AI need?' but the button says 'Book a Demo' — the action doesn't match the prompt, muddying what submitting does.

Value props

76Fair

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

Lantern AI’s headline sellsTurn your taste into hires.Lantern learns what exceptional looks like at your company in fifteen minutes, tests it against the market, then finds, interviews, and schedules people worth meeting.

Working

  • 'Taste into hires' flatters user judgment
  • Subhead names full pipeline: find to schedule
  • 'Fifteen minutes' sets concrete effort bar

Could improve

  • Hero image decorates, demonstrates nothing
  • Headline vague without the subhead
  • No proof numbers beyond YC badge

Mobile web

76Fair

The phone experience as a first-class citizen.

Mobile web annotations on the mobile capture123
  1. Human / LLM toggle — Cryptic desktop control eats header space on mobile and confuses visitors instead of collapsing into a menu.
  2. Header Book a Demo — Duplicates the form's Book a Demo button in the same viewport, splitting the tap between two identical CTAs.
  3. Placeholder demo input — The 'Who do Lantern AI need?' field reads as a required step before the button, adding friction to a demo tap it doesn't need.