MirrorsYC F26 landing page report.
← Back to YC F26

Your landing page’s weakest score is Trust.

Fix Trust

Score over time

908070605040Zep 71 on Apr 22 (Wayback)Zep 70 on May 4 (Wayback)Zep 70 on May 13 (Wayback)Zep 47 on Jun 11 (Wayback)Zep 79 on Jun 30 (Wayback)Zep 78 on Jul 14 (Wayback)Zep 78 on Jul 28 (Live)Mem0 75 on Mar 7 (Wayback)Mem0 66 on Mar 20 (Wayback)Mem0 70 on Apr 12 (Wayback)Mem0 73 on Apr 22 (Wayback)Mem0 72 on May 15 (Wayback)Mem0 70 on Jun 10 (Wayback)Mem0 78 on Jul 28 (Live)Letta 63 on Mar 7 (Wayback)Letta 59 on Jun 10 (Wayback)Letta 48 on Jun 26 (Wayback)Letta 50 on Jul 12 (Wayback)Letta 52 on Jul 28 (Live)Letta 51 on Aug 8 (Wayback)Zetta (Supermemory-style) Supermemory 72 on May 4 (Wayback)Zetta (Supermemory-style) Supermemory 70 on May 19 (Wayback)Zetta (Supermemory-style) Supermemory 75 on Jun 1 (Wayback)Zetta (Supermemory-style) Supermemory 74 on Jun 14 (Wayback)Zetta (Supermemory-style) Supermemory 75 on Jun 27 (Wayback)Zetta (Supermemory-style) Supermemory 73 on Jul 18 (Wayback)Zetta (Supermemory-style) Supermemory 75 on Jul 28 (Live)Mirrors 77 on Jul 6 (Wayback)Mirrors 73 on Aug 1 (Wayback)Mirrors 69 on Sep 20 (Live)Jul 6Aug 1Sep 20
Past reportsShow past reportsHide past reports
  1. Sep 20, 202669LiveOverall −4 · Trust −14 · Information hierarchy +10Viewing
  2. Aug 1, 202673Overall −4 · CTA −8 · Information hierarchy −6View →
  3. Jul 6, 202677First tracked report — strongest Information hierarchy 82, weakest Trust 58View →

Trust

44Poor

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

Mirrors
Zep
Mem0
Zetta (Supermemory-style) Supermemory
Trust read
Mirrors"Backed by Y Combinator" badge above the headline
Zep"TRUSTED BY AI TEAMS" logo wall with Samsung, Zscaler, HoneYBOOK, Quorum, and AGI, Inc.
Mem0GitHub star badge showing "61,868" stars in the nav
LettaResearch listing "Memory Models: Towards Agents That Learn — JUN 2026"
Zetta (Supermemory-style) Supermemory"Used by the best teams" logo strip (Caret, Jinso, Moonage, Endearing, ChatNow)

CTA

68Fair

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

Mirrors
Zetta (Supermemory-style) Supermemory
Zep
Mem0

Problems

  • click feels like a meeting, not value
  • four words, over the 3-word ideal
  • no payoff named in the label

Value props

72Fair

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

Mirrors’s headline sellsDon't let the same agent mistake reach another customer.You've already told your agent what went wrong. We're building Mirrors to carry that correction into its next decision, even in a long-running task.

Working

  • Names a specific, felt pain: repeat agent mistakes
  • Subhead explains the mechanism plainly
  • "We're building" sets honest early-stage framing

Could improve

  • No numbers or measurable outcome anywhere
  • Hero art decorates, doesn't demonstrate
  • Proof limited to YC badge and vague partner chip

Mobile web

76Fair

The phone experience as a first-class citizen.

Mobile web annotations on the mobile capture12
  1. Header CTA overlap — The nav "Talk to the founders" button overlaps the mirrors wordmark and duplicates the hero CTA below.
  2. Hero CTA width — The primary button sits at roughly half the column width instead of a near-full-bleed thumb target.

Information hierarchy

86Strong

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

Information hierarchy annotations on the desktop capture123
  1. Duplicate header CTA — Styled identically to the hero's "Talk to the founders" button, splitting one action into two equal-weight targets.
  2. Four-link nav row — "How it works", "What we're building", "Design partners", and "FAQ" are four exits competing with the single conversion.
  3. Design partners strip — "Early design partners" link creeps into the fold as another same-weight exit below the hero CTA.

Competitors

inferred from product, audience, and value prop