RoarkYC W25 landing page report.
← Back to YC W25
Roarkhttps://roark.ai/

Your landing page’s weakest score is Trust.

How Roark ranks

  1. 1.Roark77
  2. 2.Bluejay74
  3. 3.Hamming AI72
  4. 4.Coval68
  5. 5.Cekura68
  6. 6.Fixa66
Fix Trust
77Growth scoreupdated

category scores

Trust62Mobile web72Information hierarchy78CTA86Value props86
category scoresShowHide
Trust62Mobile web72Information hierarchy78CTA86Value props86

Score over time

100806040200Bluejay 52 on Apr 9 (Wayback)Bluejay 70 on May 17 (Wayback)Bluejay 66 on Jun 12 (Wayback)Bluejay 74 on Sep 21 (Live)Coval 62 on Apr 20 (Wayback)Coval 63 on May 16 (Wayback)Coval 68 on Sep 20 (Live)Hamming AI 67 on Apr 1 (Wayback)Hamming AI 67 on Apr 11 (Wayback)Hamming AI 66 on May 6 (Wayback)Hamming AI 77 on Jun 27 (Wayback)Hamming AI 76 on Aug 9 (Wayback)Hamming AI 72 on Sep 21 (Live)Cekura 67 on Apr 8 (Wayback)Cekura 70 on May 3 (Wayback)Cekura 68 on May 10 (Wayback)Cekura 66 on Jun 4 (Wayback)Cekura 66 on Jul 26 (Wayback)Cekura 68 on Sep 21 (Live)Fixa 59 on May 23 (Wayback)Fixa 66 on Sep 21 (Live)Roark 23 on Apr 17 (Wayback)Roark 29 on May 13 (Wayback)Roark 31 on Jun 12 (Wayback)Roark 74 on Jul 13 (Wayback)Roark 77 on Sep 21 (Live)Apr 17May 13Jun 12Jul 13Sep 21
Past reportsShow past reportsHide past reports
  1. Sep 21, 202677LiveOverall +3 · CTA +18 · Information hierarchy −4Viewing
  2. Jul 13, 202674Overall +43 · Value props +58 · Information hierarchy +50View →
  3. Jun 12, 202631Overall +2 · Trust +7 · Information hierarchy +4View →
  4. May 13, 202629Overall +6 · CTA +23 · Information hierarchy +6View →
  5. Apr 17, 202623First tracked report — strongest Value props 28, weakest CTA 15View →

Trust

62Fair

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

Roark
Bluejay
Hamming AI
Trust read
RoarkSOC 2 TYPE II | HIPAA BAA | BACKED BY YC badge row beneath the CTA
BluejayCustomer logo bar featuring Google, DoorDash, Navan, Abridge, Tekion, Presto, Hippocratic AI, 11x, Lendflow, Aurelian
Coval"Coval raises $28M Series A to make voice AI deployment-ready" funding banner
Hamming AI"First test report in under 10 mins • SOC 2 Type II • HIPAA (BAA available)" strip under the CTA
Cekura"New: Voice AI Orchestration Benchmarks — Retell, Vapi, Pipecat, LiveKit & more" announcement bar
FixaLive prompt composer with "Import from GitHub" and "Connect MCP tools" actions showing real product UI

Mobile web

72Fair

The phone experience as a first-class citizen.

Mobile web annotations on the mobile capture123
  1. Six-line subhead paragraph — Desktop-length copy fills half the viewport and shoves the primary CTA to the fold's bottom edge.
  2. Split CTA row — "Get started free" is squeezed to half-width beside "Book a demo" instead of a full-bleed thumb target.
  3. Four-line wrapped headline — The headline wraps into four lines on mobile, stretching the fold and delaying the value scan.

Information hierarchy

78Fair

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

Information hierarchy annotations on the desktop capture1234
  1. Header Get started button — Duplicates the hero 'Get started free' at equal weight, splitting the primary action across two targets.
  2. Five-item nav row — Product, Industries, Resources, Company, Pricing plus Sign in offer six exits before the visitor reaches the headline.
  3. Hero supporting paragraph — Five clauses of positioning, mechanics, and '500+ audio-native metrics' in one block — too dense to scan.
  4. Book a demo button — Styled nearly as prominently as 'Get started free' and repeated in the header, diluting the single conversion path.

CTA

86Strong

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

Roark

Problems

  • nav says 'Get started', drops 'free'
  • generic verb; no header echo

Value props

86Strong

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

Roark’s headline sellsTest, monitor, and improve the voice AI agents you already run.Roark is the QA and evals platform for voice AI, not another agent builder. Simulate your agent against hundreds of callers before launch, then score every production call on 500+ audio-native metrics: catch what breaks, prove the fix, and

Working

  • "already run" targets exact buyer
  • "not another agent builder" anti-positioning
  • 500+ metrics, 71→94 demo shows real product

Could improve

  • H1 lists verbs, names no outcome
  • No customer proof or result stat
  • Emotional payoff buried at subhead end

Competitors

inferred from product, audience, and value prop
#1Roark77
#2Bluejaygetbluejay.ai74 #3Hamming AIhamming.ai72 #4Covalcoval.dev68 #5Cekuracekura.ai68 #6Fixafixa.dev66