Website Audit · withbear.app

BEAR: Audit & Go-To-Market Plan

A full-surface review of BEAR’s pre-launch waitlist site — security posture, technical hygiene, design/UX, and copy — followed by a phased GTM plan, pricing strategy, and growth metrics for the waitlist-to-launch window.

Reviewed July 2026Stage Pre-launch, waitlist gatedStack Next.js on Vercel, PostHog analytics

01 · Overview

Where BEAR stands

BEAR is an AI trading coach — it reads a trader’s logged history, names costly behavioral patterns, and coaches in one voice (“Coach Bear”). Explicitly read-only and non-custodial: it never places trades. The site itself is a single-page marketing site in waitlist mode; the product is gated behind auth and not yet live to the public.

The positioning is genuinely strong.“Game film for traders” and “a coach with the receipts” are sharp, ownable lines. The read-only, non-custodial framing works as both a trust argument and a legal shield.

The site’s execution has a layout bug in the hero, a 6.3-second load time, thin technical hardening for a product that will eventually touch brokerage data, and a pricing ladder priced below what the market already pays for less — all fixable before the waitlist converts to launch.

02 · Security & Performance

Measured, not guessed

Assessed via response headers, TLS configuration, route probing, and a real Lighthouse run — not visual judgment.

60
Performance
94
Accessibility
100
Best Practices
100
SEO Structure
Bug

LCP is 6.3s — nearly 3x the “good” threshold

Largest Contentful Paint measured at 6.3s (target under 2.5s). The hero autoplay video is the near-certain culprit. A trader on a phone at 11pm who sees nothing for 6 seconds has already left.
Moderate

No Content-Security-Policy header

HSTS, X-Content-Type-Options, and X-Frame-Options are present, but no CSP — the standard next layer against injected scripts, worth having before brokerage data enters the picture.
Low

Viewport zoom is disabled

user-scalable=no blocks pinch-to-zoom — the one real accessibility failure Lighthouse caught, and easy to remove.
Note

TLS and transport are solid

Valid wildcard cert, HSTS with a 2-year max-age, HTTPS enforced throughout. No issues found.

03 · Design & UX

What the screen actually shows

Reviewed live at desktop and mobile viewports (390×844).

BEAR homepage hero, desktop
Desktop hero — stat cards overlap the subhead
BEAR homepage hero, mobile
Mobile hero — clean, no overlap at this size
Bug

Hero stat cards overlap the subheadline (desktop only)

Confirmed via Lighthouse + live rendering: the overlap bug only exists at desktop widths. Mobile hides the floating cards entirely and renders clean — scope the fix accordingly.
Moderate

Pricing sits ahead of proof

The pricing ladder is fully built out before any visitor can use the product, with only three unattributed testimonials as proof.
Note

What’s working

The waitlist page (bear’s eyes, single field, one CTA) is the best-executed screen on the site, on any device — the homepage hero should learn from its restraint.

04 · Copy & Positioning

The claim is buried in paragraph three

“You size up after two wins. That’s your pattern, and it’s the most expensive one you have.”

This is the strongest copy on the site — specific, evidence-based, slightly uncomfortable. It does more selling than any feature bullet.

The preemptive claim is buried.“Bear reads; Bear never trades for you” appears four times across the page but never as the load-bearing headline. It’s the entire wedge against journals and signal services — it should be the first sentence, not the third section.

05 · GTM Plan

Fix, then prove, then grow

Four phases, sequenced so proof exists before spend.

Weeks 1–2

Fix the trust surface

Hero bug, testimonial attribution, security headers, H1 rewrite — before a single paid visitor arrives.

Weeks 2–5

Seed proof before spend

Recruit 20–30 beta traders directly, ship 3–5 named case studies, replace the medallion with a real demo.

Weeks 5–8

Open the funnel

Organic-first on FinTwit, small trading-educator partnerships, flip off noindex, add real waitlist scarcity.

Weeks 8–12

Launch to the waitlist

Cohorted invites (100–200 at a time), Founding tier sold to the proven beta cohort, not cold traffic.

Claude prompt · Fix the hero

In the homepage hero, fix three things:
1. The floating stat cards ("Risk management +18% improved", "Win rate +15%") overlap the subheadline text on desktop. Fix the layout so they never overlap — anchor them to flank the medallion graphic only, or move them below the subhead.
2. Rewrite the H1 to lead with the fact that Bear reads every trade but never places one — that's the one claim no competing journal or signal service can make. Give me 3 short options.
3. Replace the three testimonials with a component that requires a real first name and a verifiable detail (e.g. "247 trades logged"), instead of unattributed quotes.

Claude prompt · Fix load speed

The homepage's Largest Contentful Paint is 6.3 seconds, measured with Lighthouse — the hero autoplay video is the likely cause. Fix this:
1. Add a poster image so the hero shows something instantly instead of waiting on the video.
2. Compress the video file, or lazy-load it so it starts after the page is interactive instead of blocking first paint.
3. Re-run Lighthouse after the fix and confirm LCP is under 2.5s.

06 · Moat & Virality

Trading DNA is a viral loop sitting unused

BEAR already computes a shareable, named archetype ("The Momentum Chaser") and a Trader Score — structurally identical to Spotify Wrapped or Wordle. It is currently buried as feature #5 of 8.

Ship a shareable, branded result card that generates automatically after a user’s first 10 logged trades, with a one-tap share to X. Every share is a free, high-trust impression in front of an audience that already self-selects as traders — the cheapest CAC channel available, because the poster is vouching for the product themselves.

Claude prompt · Trading DNA share card

Build a shareable "Trading DNA" result card. Requirements:
1. Auto-generate it the first time a user hits 10 logged trades.
2. Show their archetype name (e.g. "The Momentum Chaser"), their Trader Score, and one real specific stat from their own history — no generic praise.
3. Dark background matching our site's palette, sized for a clean X/Twitter share preview (1200x675).
4. Add a one-tap "Share to X" button that pre-fills a tweet with the card attached, plus a "Copy image" fallback.
5. No dollar amounts or account info on the card by default — let the user opt in if they want to add a P&L stat.
6. Track "card generated" and "card shared" events in PostHog.
LayerWeak (copyable in a quarter)Strong (compounds)
Personalization“AI reads your trades”Coaching gets sharper the longer a trader stays
Category“AI trading coach” (generic)Own “game film for traders” as the category name
Trust“We’re read-only” (a claim)Read-only architecture is expensive for an incumbent to retrofit
Switching costNone todayMonths of logged “game film” becomes an archive worth keeping

07 · Pricing

Priced like a commodity, written like a premium coach

Established journals with less capability already charge more than BEAR's mid tier: Tradervue and TraderSync run up to $79.99/mo for a plain journal — no coaching layer at all.

Free

$0

  • Unlimited trade logging
  • Trading DNA share card
  • ~5 Coach Bear msgs/mo

Replaces the confusing 7-day trial.

Recommended

Pro

$24–29/mo

was $10.99–20.99

  • Unlimited Coach Bear
  • Full behavioral analytics
  • Weekly reports

Priced at parity with Tradervue/TraderSync — a coach is a bigger claim than a journal.

Founding

$59–79/mo

was $99.99

  • Everything in Pro, for life
  • Real capped spots, shown live
  • Direct line to the founder

Sell to the warm beta cohort, not cold visitors.

Claude prompt · Fix pricing tiers

Fix our pricing page and tiers:
1. Remove the free trial that currently converts to the same price as our top paid tier — replace it with a real free plan (unlimited trade logging, capped Coach Bear messages, no expiry).
2. Update the main paid tier's price to $24–29/month.
3. Keep the Founding tier, but cap it at a real number (e.g. 250 spots) and show the remaining count live on the page instead of just claiming spots are limited.

08 · Growth Metrics

What to watch for the first 1,000 users

CAC and LTV are premature at this scale. Watch whether the core loop works, and whether it spreads.

Visitor → waitlist signup ratetarget 15–25%Cheapest, fastest read on whether the hero and copy convert.
Activation ratesignup → first trade loggedTests the real product promise, not just curiosity.
Week-1 / Week-4 retentionof activated users onlyTells you if this is a habit or a novelty.
Referral k-factor>0.3 is a real signalTells you if the referral loop is doing any work.
Sean Ellis PMF score≥40% “very disappointed”The standard early-stage signal — works at low N.

Your site

Want a report like this on your site?

Get your free diagnosis