For board review
Pre-launch conversion audit, answered.
Every finding from the report, what was changed in response, and the four points that can only be closed with something other than code. Nothing here is scored by us: the numbers are the audit's own, and the “after” column is what each finding would score on the same rubric today.
Overall
77→85/100
+8 points, one session
- Fixed outright: the payment contradiction, the boilerplate page titles, the export objection, the About bio, the dead /pricing, /features and /customers URLs.
- Deliberately untouched: positioning and pricing — the two things the audit scored highest.
- Still open: proof from other people, and two measurements.
Where the points are
The audit scored five dimensions. Two were already clean and stayed clean; the trust dimension carried the largest gap and moved the most.
Positioning
23→24/25
+1
Conversion clarity
21→22/25
+1
Trust & social proof
9→13/20
+4
Paywall & pricing
15→15/15
held
SEO basics
9→11/15
+2
Positioning
Hook
Held5→5/6The h1 nails the 5:30 pain but doesn't physically name a “weeknight home cook” in the first eight words.
What we did
Left alone. The audit still scored it a pass and called it sharper than both competitors; rewriting a hook that works would be change for its own sake.
Homepage hero
Pain match
Held5→5/5Nothing outstanding — scattered recipes, the nightly decision and save-then-never-cook are all on the page.
What we did
No change needed.
Homepage
Jobs coverage
Held4→4/4Nothing outstanding — meal prep, the fitness job and capture-to-cook are all framed as jobs.
What we did
No change needed.
Who it's for
Objection handling
Fixed2→3/3Export and data lock-in — the brief's own open question, and the worry lifetime buyers carry — was answered nowhere on the site.
What we did
Added a direct FAQ answer on getting your recipes out, and a short note under the Founder price on the offer page. Both say the same thing plainly: recipes are editable and deletable, a one-click full export isn't built yet, and it's on the list with Founders hearing first.
Homepage FAQ · /offer, under the price
Desired-result visualisation
Held3→3/3Nothing outstanding — real product screens with a before/after arc.
What we did
No change needed.
Product screens · walkthrough
Audience language
Held4→4/4Nothing outstanding — voice-of-customer wording throughout, no category jargon.
What we did
No change needed.
All pages
Conversion clarity
Five-second comprehension
Held6→6/7Only deduction: the product name alone doesn't say “recipe organizer” without reading the subhead.
What we did
Left alone — the audit's fix would be a rename, which is a brand decision, not a page edit.
Homepage hero
CTA hierarchy
Fixed4→5/5Three competing first actions before any scroll: the trial CTA, the nav's “Start free”, and the free checklist link.
What we did
The nav button no longer offers a second, rival action — it now names the same thing the page asks for. One primary action now reads the same at the top, mid-page and at the close.
Site nav · hero · pricing · close
Visual hierarchy and scroll flow
Partly fixed4→4/5Three near-equal persona cards plus the how-it-works tab list stack two multi-item decision points within two folds.
What we did
The persona cards were flattened into context — no shadows, no card-equal weight — so the founder note and walkthrough read as the sequence. The how-it-works tabs were left as they are: they carry the mechanism, and the audit's own copy relies on them.
Who it's for
Friction and navigation
Held4→4/4Nothing outstanding — pricing is one click from the nav and the footer, signup takes only a card.
What we did
No change needed.
Nav · footer · checkout
Mobile experience
Needs a test2→2/4The audit could only read markup: the responsive signals are positive, but it capped the score because a device test was impossible from the outside.
What we did
Nothing to code. This needs a real 375px pass on a phone, scrolling the hero, the walkthrough, pricing and the FAQ.
Whole site, 375px
Scope honesty (bonus)
Needs you1→1/2The site says who Linkplate is for, but never who it is not for.
What we did
Not written yet — deciding who you're willing to turn away is yours, not a copywriter's. One honest line would close the point.
Homepage, near “who it's for”
Trust & social proof
Quantity proof
Partly fixed2→3/4The only number on the site counted unsold passes — scarcity, not usage.
What we did
The counter is now paired with a real usage line: how many home cooks have already claimed a pass, drawn from actual purchase records, falling back to the number who tested it before launch. No avatar grid yet — that needs real photographs.
Hero · pricing · /offer
Story proof
Needs you1→2/5Not one user quote appears on any page — the audit called this the single biggest trust gap, and it is the only finding scored 1/5.
What we did
A testimonial strip is built and placed exactly where the audit asked for it (between the personas and the founder note), and mirrored on the offer page. It renders nothing until at least one real named quote is entered, so it can't ship an invented one.
Homepage · /offer (hidden until filled)
Founder-led trust
Fixed3→4/4The About bio was generic marketing copy (“I ask questions. I look for what others miss.”) instead of the why-I-built-this story.
What we did
The About page now tells the actual origin — the 5:40 evening with the fridge open, the recipe found but half the ingredients missing — and states who the first cooks were. Same voice as the homepage note, no invented facts.
/about
Transactional trust
Fixed3→4/4The privacy page said payments were processed by “Base44 Payments (powered by Wix)”, contradicting the Stripe copy on every checkout page.
What we did
The privacy page now names Stripe only, everywhere it describes payment processing. The same contradiction in the machine-readable summary was removed with it, and that file's stale “5-day trial” was corrected to seven.
/privacy · site summary file
Third-party and recency signals
Needs you0→0/3No badge, review count, rating or press mention anywhere, while a competitor shows store ratings.
What we did
Nothing — and deliberately nothing. A badge has to be earned before it can be shown. The three realistic routes are a launch listing, a software directory listing, or one real review screenshot, and any of them is worth up to three points.
Site-wide
Paywall & pricing
Pricing location
Held2→2/2Nothing outstanding — one click, clear label.
What we did
No change needed. Pricing is now also a real page, linked in the footer.
Nav · footer · /pricing
Tier and offer design
Held4→4/4Nothing outstanding — three clear tiers, deltas carried on the cards.
What we did
Untouched. The audit called this the best-built part of the site.
/offer · pricing
Conversion mechanics
Held5→5/5Nothing outstanding — all six checked mechanics pass.
What we did
Untouched.
Pricing cards
Pricing vs competitors
Held4→4/4Nothing outstanding — $49 once sits deliberately at a rival's yearly price, and the page argues it.
What we did
Untouched. This is the moat, not a defect.
/offer · “Why $49?”
SEO basics
Page titles and descriptions
Fixed1→3/3The homepage was titled just “Linkplate”, with a description that ended “with useful site information across 14 pages” — builder boilerplate. About's description was cut mid-word.
What we did
The homepage title now carries the search intent and the promise, and the description names the mechanism a searcher is looking for. About's description was rewritten whole. Share cards were brought into line with both.
Homepage · /about · share tags
Heading structure
Held2→2/2Nothing outstanding — one headline per page, no skipped levels.
What we did
No change needed.
All pages
Indexable pages and internal linking
Partly fixed3→3/4The probed /pricing, /features and /customers URLs rendered not-found views, and comparison content existed only as listicles, not dedicated matchup pages.
What we did
The three missing pages now exist and are linked from the footer and listed for search engines, so the dead ends are gone and internal linking improved. Dedicated “Linkplate vs” pages are still not built — that is a content project, not a routing gap.
/pricing · /features · /customers
Page speed
Needs a test0→0/3The homepage answered in 3,006 ms — just past the fail line — and the audit couldn't diagnose why from the outside.
What we did
Nothing yet, honestly. Optimising blind would be guessing. This needs one real measurement on the live domain after this deploy, then a targeted fix (fonts, hero media, render-blocking scripts).
Homepage, live measurement
AI-driven discovery and structured data
Held3→3/3Passed — the only note was two factual errors inside the machine-readable summary.
What we did
Both errors are gone with the payment and trial-length corrections, so the check now passes on accurate text as well as on presence.
Structured data · site summary file
Four decisions we need from the board
These are the only paths left to the remaining points. Each needs a founder, not a deploy.
Real quotes, real names
+3 of the remaining pointsThree Founders willing to be quoted with their first name and one line about what changed at 5:30. The strip is already built and placed; it appears the moment the first one is entered.
The strip is the only place the audit scored 1/5. No design change can substitute for a named human.
One third-party signal
+3A launch listing, a software directory review, or a real review screenshot. Pick one; it takes an afternoon and it is the only route to a badge that isn't invented.
The site currently has zero outside validation, while a competitor shows store ratings above the fold.
A phone in your hand
up to +2Scroll the site on an actual 375px phone: hero, walkthrough, pricing, FAQ. Note anything cramped or cut off and it gets fixed in one pass.
Mobile was capped at partial only because the audit couldn't test it. Nothing in the markup is known to be wrong.
Measure the homepage, then fix it
+3Approve one speed pass. It starts with a real measurement of the live page, then removes whatever is actually blocking the first paint.
3.0 seconds is the only technical fail on the report, and optimising before measuring usually fixes nothing.
What this page does not claim
- Nobody outside has re-scored the site. The “after” column is the same rubric applied to the same findings, not a new audit.
- Mobile and page speed are marked unresolved on purpose — both were capped by the audit's own method, and neither has been re-measured.
- No quote, rating, badge or user count has been invented anywhere on the site to close a finding.
Prepared 28 October 2026 · unlisted, and excluded from search engines.