Why the redesigns beat beefbootlegger.com — the full analysis

Every claim below was verified against the actual code by three independent AI reviewers — Kimi K3 (read all four files + the audit), Codex (adversarial code review, 100k+ tokens of checking), and Sakana Fugu Ultra (business strategy). Where they found problems in OUR concepts, we fixed them and logged the fix. This is the receipts page.

Old site: WordPress + WooCommerce 4 concepts: A Speakeasy · B Butcher's Counter · C Meat Club · D The Vault (3D) Rulebook: Drey's UI-STANDARD

1. The scoreboard — 16 documented problems on the live site

We scanned every page of the live site and documented 16 concrete problems. Kimi and Codex then independently checked our concepts' markup to verify each one is actually fixed — not assumed fixed. Verdicts below are theirs, updated after our fix round.

Problem on the live site todayABCD
Two competing Login/Register menus (custom pages + wp-login)FIXEDFIXEDFIXEDFIXED
Broken phone link — href="http://(727)-472-5456" is an invalid URLFIXEDFIXEDFIXEDFIXED
Footer year missing in the pre-JavaScript markup (appears once scripts run — crawlers see the bare "Copyright ©")FIXEDFIXEDFIXEDFIXED
Stat counters read "0 Pounds of Meat / 0 Happy Customers" until JavaScript runs — the zero-state is what Google, crawlers, and slow connections see. (Verified live: in a full browser they do animate up to 6.4K / 1,028 / 7 / 82 — so we kept his real numbers and dropped the zero-risk.)FIXEDFIXEDFIXEDFIXED
Events in the nav opens a completely empty pagePARTIALFIXEDPARTIALPARTIAL
Contact page leaks WordPress blog sidebar (Recent Posts, Archives, honeypot text)FIXEDFIXEDFIXEDFIXED
Price text duplicated: "$22.00 – $150.00 Price range: $22.00 through $150.00"FIXEDFIXEDFIXEDFIXED
Raw variation tables; no grade help or reviews at the point of choicePARTIALPARTIALPARTIALPARTIAL
Flat 38-product wall mixing seafood/chicken into a beef brandFIXEDFIXEDPARTIALFIXED
Three rotating hero taglines — no single messageFIXEDFIXEDFIXEDFIXED
No "How it works"; local delivery vs 48-state bulk never explainedFIXEDFIXEDPARTIALFIXED
VIP ($7.77/mo) and Bootlegger Bucks buried in menusFIXEDFIXEDFIXEDFIXED
No social links anywhere despite active IG/TikTok/FBFIXEDFIXEDFIXEDFIXED
About page repeats counters + a second testimonial block, empty headingsFIXEDFIXEDFIXEDFIXED
Two competing "reviews" experiences, both uncurated wallsFIXEDFIXEDFIXEDFIXED
Trust signals scattered — no guarantee, no delivery-zone storyPARTIALPARTIALPARTIALPARTIAL

Honest reading of the PARTIALs: these are homepage concepts, not full sites. "Point-of-choice grade help" needs real product pages, "Events" needs a real events page with dates, and "trust" wants a delivery-zone map — all scoped for the production build, impossible to fake honestly in a one-page concept. The current site fixes NONE of the 16.

Credit where due — what the live site does BETTER (verified by rendering it, then confirmed by Kimi K3)

And the animations question, directly: his animations are a 3-slide tagline rotator, count-up counters, and stock fade-ins riding on jQuery + 20 script files (4.4MB page). The rotator IS the "no single message" problem, and the counters ARE the zero-state SEO problem — the two flashiest animations on the site are the two causing defects. Ours animate with intent (entrance reveals, tilt, and D's draggable 3D vault), weigh a fraction as much, and every one respects the reduced-motion accessibility setting, which stock jQuery counters and sliders do not.

2. Scored against the house rulebook (UI-STANDARD)

Drey's UI-STANDARD is the design rulebook every client build gets judged by. Here's the old site vs the concepts on its core rules, scored /10:

Rule 1 — "Kill the vibe-coded tells" (one message, no icon spam, a real layout decision)

Current site
2
A — Speakeasy
9
B — Butcher's Counter
8
C — Meat Club
9
D — The Vault (3D)
9

Old site: 3 rotating taglines, decorative "shape" PNGs, identical card grids — the swap-the-logo test fails instantly. Each concept has one message, one accent color, a distinctive layout move (dot-leader menu / occasion picker / member table / 3D vault), and passes the litmus: swap the logo and it still looks like nobody else's site.

Rule 2 — Motion as a craft signal (with prefers-reduced-motion respected)

Current site
2
A / B / C
8
D — The Vault (3D)
9.5

Old site's only "motion" is the tagline rotator and counters that visibly break. Every concept ships intentional entrances, hovers, and one signature moment — and every animation is wrapped in a reduced-motion check, including smooth-scroll (Codex caught that we'd missed smooth-scroll; fixed in all four).

Rules 5 + 6 — Real depth, real interaction hygiene

Current site
3
A / B / C / D
8

Concepts: layered shadows on cards that should feel like objects, SVG stars instead of icon-font spam, focus-visible outlines on every interactive element, skip-to-content links, per-page tab titles + meta descriptions, semantic header/nav/main/footer. Old site: flat sections, default WooCommerce chrome, honeypot text visible to humans.

3. The conversion math — from Instagram to add-to-cart

Kimi traced the first-purchase path for a cold visitor landing from Instagram (where Stu's traffic actually comes from):

SiteSteps to a buyable productWhat kills the sale along the way
Current site5+ stepsShop nav → scroll a 38-item wall across 2 pages → open a product → decode a raw variation table with duplicated price text → add to cart. No grade help, zero reviews shown, delivery area unexplained.
A — Speakeasy2 stepsGorgeous, but the menu is a price list — production build must make rows clickable.
B — Butcher's Counter2 stepsOccasion chip → 3 recommended cards with price + add-to-cart. Kimi's pick for cold traffic: answers "what is this / can I trust it / what do I buy" in the first viewport.
C — Meat Club2 stepsLeads with a $250 box + membership — a big first ask 10 seconds off a reel.
D — The Vault2 stepsThe 3D vault is the product browser — spin, stop on a cut, price is right there. Needs the same cart wiring as the others.

4. What each fix is worth in dollars (Fugu Ultra's ranking)

Fugu Ultra ranked all ten shared fixes by expected revenue impact for THIS business — a relationship-driven local brand where one recovered bulk order is worth $1,550–$5,700:

  1. Delivery clarity — the #1 leak. Buyers can't tell Tampa hand-delivery from 48-state bulk shipping today; confused carts become cancellations and refunds. One recovered bulk order pays for the whole redesign.
  2. Clean pricing — "Price range: $22.00 through $150.00" reads like hidden costs; corporate hosts can't budget against it. Exact prices raise product-to-cart conversion.
  3. Working contact + guarantee — four-figure buyers call before they buy. A broken phone link on a $5,700 decision is a closed door.
  4. Mobile-first — his traffic is Instagram/TikTok; that lands on phones. (This finding is also why we added phone-visible mobile nav after review.)
  5. Single message — one static proposition instead of a 3-slide roulette.
  6. Curated reviews at decision points — moves Stu's offline reputation online, where the premium price needs it.
  7. Grade education — kills choice paralysis and creates credible Prime upsells.
  8. Surfaced VIP/Bucks — AFTER an economics audit (see the warning below).
  9. One login — fewer resets, faster reorders.
  10. Real numbers — "0 Pounds of Meat" makes the business look dead. Cheapest fix, immediate trust.

The money finding nobody asked for: VIP + Bucks stacking

Fugu Ultra ran Stu's own numbers: if Bootlegger Bucks (a 9.09% effective discount) stacks on top of VIP's 10%, a $100 product effectively sells for ~$81.82 — an 18.2% discount before the membership fee. And a VIP buying a $5,700 whole beef takes a $570 discount against $7.77 of monthly revenue. Before ANY redesign promotes these offers harder, Stu needs stacking rules, bulk exclusions, and caps in writing. We already removed every "stacks with VIP" claim from the concepts. This one insight may be worth more than the website.

5. The adversarial pass — what the reviewers caught in OUR work, and what we did

We didn't just grade the old site — we had all three models attack the new concepts. Every legitimate hit got fixed the same hour:

Catch (who)Status
Mobile nav was deleted below 900px — nav AND phone number vanished on the exact devices his Instagram traffic uses (Kimi + Codex)FIXED — phone number stays visible on mobile in all 4 concepts; full hamburger menu specced for production
"NEVER FROZEN" hero sticker contradicted the frozen bulk-shipping program (Kimi + Codex, independently)FIXED — now "CUT TO ORDER"
"No antibiotics/hormones" was documented for the Grass line only, but we'd claimed it brand-wide (Codex)FIXED — scoped to "Grass-Fed Line" in every concept
Invented offer terms: Bucks "never expire", "stack with VIP", "cancel anytime"; "unmarked" deliveries; "aged" beef (Codex)FIXED — every undocumented claim removed or softened to what reviews actually say
Contrast failures in C: guest-column text 2.65:1, red kickers 3.82:1, white-on-red buttons 4.31:1 (Codex, measured)FIXED — guest text lightened, kickers/footers moved to darker red or light coral, buttons darkened to pass 4.5:1
Smooth-scroll not gated by prefers-reduced-motion; no skip links; C's comparison table had no overflow wrapper or label (Codex)FIXED in all 4 — gated scroll, skip-to-content links, scrollable labeled table
Occasion chips lacked aria-controls (Codex)FIXED
Reveal animations left the page blank if JavaScript ever failed (found in our own testing)FIXED — content is visible by default; animation only activates when JS confirms it's running
"My grill started clapping — Julian B." flagged as invented (Codex)FALSE POSITIVE — it's a real Google review from the live-site scrape; Codex only saw the audit summary. Kept, and flagged for quote-verification with Stu anyway.
Full-size 2100px images used in card-size slots; no width/height attributes; no srcset (Codex)ACCEPTED for local concept files (instant off disk) — resized/responsive images are line 1 of the production checklist
No cart, search, account, or ZIP-check — nothing is actually purchasable (Kimi)BY DESIGN — these are direction concepts. The real build (Shopify) brings cart/checkout/accounts natively; ZIP-gate is specced in the README.

6. Concept D — what the 3D version is, and the rules that keep it honest

What moves (product evidence)

  • The Vault — a real 3D rotating cylinder of his eight core cuts. Drag to spin, buttons to step, front card always stops readable with name + price.
  • Hero depth — the ribeye photo, the light, and the headline sit on separate 3D layers that drift gently with the mouse, then settle.
  • Ember particles — a lightweight canvas flame-spark layer over the hero. No libraries, no WebGL dependency.
  • Grade cards — tilt toward the cursor like physical cards.

What never moves (the buying interface)

  • Prices, delivery terms, the local-vs-bulk split, reviews, guarantee, contact, and every button — static, always readable.
  • Normal scrolling — no scroll-jacking, no forced tours.
  • Reduced-motion users get a fully static page; no-JS users get a flat product grid instead of the vault. Nothing breaks.
  • Phones: touch-swipe spins the vault; parallax and tilt switch off.

Fugu Ultra's guardrail, followed to the letter: "Move product evidence, not the purchasing interface." And its success metric if Stu picks D: non-refunded revenue per visitor, A/B tested against a static version — spin-time and scroll depth are diagnostics, not proof. Honest risk (also Fugu): D is the most expensive concept to maintain and the easiest to get wrong on cheap phones — it should A/B its way in, not launch on faith.

7. The verdicts

Kimi K3 — full-corpus review

Concept B converts cold Instagram traffic best. "It answers the three cold-traffic questions in the first viewport... its skeleton is the shortest honest path from curiosity to checkout."

Codex — adversarial code review

Concept B is the strongest first-client demo — "occasion-based discovery, grade comparison, delivery split, and restrained voice make the business easiest to understand." A has the cleanest performance + brand fit; C was weakest pre-fixes.

Fugu Ultra — business strategy

B is the safest commercial base for families AND corporate hosts — with the warning that clarity can commoditize: keep Stu, hand-delivery, and Tampa identity central, or B becomes any online butcher.

The recommendation (one, not a menu)

Build on B's skeleton, dressed in A/D's skin, with C's membership table as its own page. All three reviewers independently picked B's structure as the best seller — and all three warned its plainness is the risk. So: B's occasion picker, grade table, and delivery split as the layout; the Speakeasy/Vault dark-premium look and voice as the brand layer; the member-vs-guest table from C on a dedicated VIP page. The Vault carousel from D becomes the signature "wow" module on the homepage — A/B tested, per Fugu. Confidence: high on the structure call (3-for-3 reviewer consensus), medium on the exact visual blend until Stu reacts to A vs D.

The production sitemap (the answer to "his site has tabs")

Kimi K3's page architecture for the Shopify build — the concepts are homepage directions, the real site is six pages: Home (hybrid homepage: stats, occasion picker, grade table, delivery split, Google reviews, Stu, VIP teaser) · Shop (full 82-product catalog with grade/cut/occasion filters) · VIP / Meat Club (C's member-vs-guest table as a revenue page) · Bulk Beef (quarter/half/whole with per-pound math) · Our Story (Stu + the 7 farms + the lore) · Contact / FAQ (delivery zones + how-it-works). Two of his current tabs die honestly: Events (an empty page today) and standalone Reviews (absorbed into Home and product pages). The redline for Stu: current — 7 tabs, one empty, VIP buried. New — 6 pages, zero dead ends.

8. What's still open (the honest list)

Prepared 2026-07-31 · Sources: full scan of beefbootlegger.com (all pages, 38 products, 122+ reviews, socials) · research/site-audit.md · UI-STANDARD.md · Reviews by Kimi K3, Codex (gpt-5.6-sol, read-only sandbox), and Sakana Fugu Ultra v1.1 · All four concepts: open the comparison page