Well built, badly positioned. The craft is real: the product screenshots show a team actually debugging together, the writing is human, and the page paints in under half a second. The problem is placement. The two lines that explain what Neatlogs is and why it is not just another observability tool sit fourteen screens down and inside a collapsed FAQ, while the hero spends its space on a generic promise and a watercolour bridge. The closing section promises free to start, then offers only a demo call. Almost all of it is copy and reordering rather than a redesign, so the score is very recoverable, starting with the hero.
Scored against 27 checks in three weighted categories,
plus 11 signals measured directly from the page. Section 08 shows how the score is built.
Section 01
Verdict
The bottom line
Your site is well built and badly positioned: the two lines that explain why Neatlogs is different are fourteen screens below where a buyer needs them.
01
The leak
Your closing section promises free to start, no credit card required, then offers one button: Book a Demo.
02
The gap
The hero never says what category this is. The phrase collaborative debugging workspace for AI agents sits in a collapsed FAQ at the bottom.
03
The imagery
A watercolour bridge holds the hero and figures fly behind a laptop in the footer. Neither shows the product, and both look generated to fill the space.
04
What is strong
The product screenshots, the page speed at 408ms, and the writing. This is a craft problem in placement, not in execution.
The first five seconds
What a stranger takes from the hero
Lands
Who it is for, and a promise: for teams shipping agents, feedback to fix, fast.
Does not land
What kind of product it is. No category, no product shot, and nothing separating you from LangSmith or Braintrust.
Where it stands
Three categories, 27 findings
The summary. Every finding is detailed below
211122
2 critical11 high12 medium2 low
UX and conversion
60/100 · At risk
The mechanics are competent and the writing is good, but the page leads with a promise instead of a position. The category, the differentiator, and the proof are all either below the fold or missing, and the closing call to action contradicts its own copy. These are placement problems, which means most of the score is recoverable without new design work.
Visual and behavioral
81/100 · Solid
Craft is high and the product screenshots are genuinely persuasive. The hierarchy is spending its best position on decoration and burying the strongest asset, and the page runs about twelve screens with pinned sections that slow a skimmer down. Reordering fixes most of this.
Compliance and trust
92/100 · Strong
The basics are in place: privacy policy, terms, glossary, and a support route are all linked. The open items are accessibility, where 31 elements fail AA contrast and 20 mobile touch targets fall under 44px, and a visitor identification tracker that fires before any consent and is currently returning errors.
Section 02
What is not working
27 findings, most severe first. Each is grounded in the captured page.
UX and conversion
12 issue(s) · scores 60/100
The hero never says what kind of product this is
hero, desktop and mobile · value prop clarity
criticalobservedquick win
Above the fold the page offers an audience and a promise but no category. The eyebrow reads FOR TEAMS SHIPPING AGENTS and the headline reads feedback to fix, fast. Neither names the product space. The supporting line says one workspace for teams to find issues, align on failures, and ship fixes together, which describes a workflow but would equally describe Linear, Sentry, or a shared Notion page. The word debugging does not appear until section two, and the phrase collaborative debugging workspace for AI agents appears only inside a collapsed FAQ item at the bottom of the page.
Evidence
Eyebrow: "FOR TEAMS SHIPPING AGENTS"
Headline: "feedback to fix, fast"
Subhead: "one workspace for teams to find issues, align on failures, and ship fixes together."
FAQ 01, collapsed by default: "Neatlogs is a collaborative debugging workspace for AI agents."
Why it matters
A buyer who cannot place the product in a category cannot decide whether it is relevant to them, so they leave rather than scroll. This is the single largest drop-off risk on the page, and it is being caused by the best sentence on the site sitting fourteen screens below where it is needed.
Fix
Promote the FAQ definition into the hero. Set the eyebrow to the category and the persona together: "COLLABORATIVE DEBUGGING FOR AI AGENTS". Keep the audience line in the subhead. The headline can stay short if the category is doing its job in the eyebrow.
The closing section promises self-serve and offers only a sales call
start debugging together section, near the bottom · cta clarity
criticalobservedquick win
The final conversion section is headed start debugging together with the reassurance free to start, no credit card required directly underneath. The only button in that section is Book a Demo. The copy removes the risk of signing up and then the interface refuses to let the visitor sign up. A reader who has scrolled the entire page, been convinced, and reached for the free option is handed a calendar instead.
Evidence
Heading: "start debugging together"
Supporting line: "free to start, no credit card required"
Only control in the section: "Book a Demo"
Why it matters
This is the highest intent moment on the page and the offer contradicts itself. Visitors who wanted to self-serve will not book a call, and the promise of free to start now reads as bait, which costs trust as well as the conversion.
Fix
Make the primary button in this section match the promise: a dark Start free button next to a quieter Book a demo. If the product genuinely requires a call before access, delete free to start, no credit card required.
The differentiator is buried in a collapsed accordion
FAQ, bottom of page · value prop clarity
highobservedquick win
The clearest statement of why Neatlogs is not just another observability tool sits in FAQ item two, which is collapsed by default at the very bottom of a page roughly twelve viewport heights tall. The second best statement, Simple enough for domain experts. Detailed enough for developers, is a subhead in section two. Neither is visible in the first screen. In a category that already contains several well funded tools, leading with an outcome alone gives the visitor nothing to compare on.
Evidence
FAQ 02, collapsed: "Standard observability tells you what happened. Neatlogs is built around the debugging workflow: shared traces, comments, evaluations, tasks, and suggested fixes, so teams can move from issue to fix faster."
Section two subhead: "Simple enough for domain experts. Detailed enough for developers."
Hero carries no comparative or differentiating claim
Why it matters
In a crowded category the buyer is not asking what does it do, they are asking why this one. Answering that below the fold means only the visitors who were already interested ever see the answer.
Fix
Move the observability contrast into the hero subhead, in the reader's terms: "Observability tells you what happened. Neatlogs is where your team works out why, and ships the fix." Then keep the FAQ answer where it is for the people who scroll.
No customer, no logo, no number anywhere on the page
whole page · trust signals
highobservedmedium
The page contains no customer logos, no named team, no testimonial, no usage figure, and no funding or backing signal. The section headed fits into the way your team works shows the tools it integrates with, which is a compatibility claim rather than evidence anyone uses it. The see it in action section offers walkthroughs described as running in real teams' workflows, but the teams are not named.
Evidence
No logo wall, testimonial, or metric found in the page copy
"fits into the way your team works" shows integrations, not customers
"short walkthroughs of neatlogs running in real teams' workflows" with no team identified
Why it matters
For infrastructure a team will route production traces through, evidence that someone else already trusts it is close to a requirement. Without it the buyer carries all the risk of going first.
Fix
Add one proof element directly under the hero CTAs, even a modest one: a row of three named teams, a trace volume figure, or one attributed quote. If nothing is publicly nameable yet, use a number, for example traces processed to date.
No pricing, so the buyer cannot qualify themselves
whole page and navigation · objection handling
highobservedmedium
There is no pricing page and no pricing link in the navigation or the footer. The only commercial signal is free to start, no credit card required in the closing section, with no indication of what happens after free, what a paid plan costs, or what triggers the upgrade. The navigation offers Docs, Careers, and Changelog but nothing about cost.
Evidence
Navigation: "Docs", "Careers", "Changelog"
Only commercial claim on the page: "free to start, no credit card required"
No pricing link found anywhere in the page
Why it matters
Cost is the first question a buyer takes to their manager. Hiding it forces every evaluation through a sales conversation, which filters out the self-serve developer who would otherwise have tried it that afternoon.
Fix
Publish a pricing page, even a provisional one with a free tier and a contact us tier, and link it from the main navigation. If pricing is genuinely undecided, state the free tier limits on the page instead.
JavaScript console errors
code · conversion friction
highobservedquick win
2 JavaScript error(s) thrown on load. Broken scripts break flows.
Evidence
Failed to load resource: the server responded with a status of 404 ()
Failed to load resource: the server responded with a status of 429 ()
Why it matters
Errors on load mean some scripts did not run. Whatever they powered is silently broken for some users.
Fix
Fix the thrown errors, then add error monitoring so regressions are caught before users hit them.
Heavy pages lose visitors before they render, and the loss is worst on the mobile connections most visitors use.
Fix
Compress and correctly size images, serve modern formats, and defer non-critical scripts.
Three different labels for what look like two destinations
navigation and hero, all viewports · cta clarity
mediumlikelyquick win
Above the fold the visitor is offered Sign up in the navigation, Book a Demo in the navigation, then Book a Demo and Get Started in the hero. Get Started and Sign up appear to lead to the same place but are named differently, and Book a Demo is presented twice within the same screen. The visitor has to work out that four controls represent two choices.
Evidence
Navigation: "Book a Demo", "Sign up"
Hero: "Book a Demo", "Get Started"
Four button-styled controls detected above the fold
Why it matters
Repeating the same option under two names makes the page feel less considered and splits clicks across duplicate paths, which makes the analytics harder to read as well as the page.
Fix
Pick one label per destination and use it everywhere. Suggested: Start free for the product, Book a demo for sales. Remove the duplicate Book a Demo from the navigation so the hero owns that choice.
The page title is one word, so search and tabs say nothing
document head · messaging clarity
mediumobservedquick win
The title element is just neatlogs. The meta description underneath it is strong and specific, and the Open Graph title is better still, but the browser tab, the search result heading, and every bookmark say only the brand name. A visitor with several tabs open has no way to identify this one, and a search result gives a stranger no reason to click.
Evidence
Title: "neatlogs"
Meta description: "From trace analysis to automated task management, neatlogs provides all the tools your team needs to identify, understand, and fix AI agent issues efficiently."
Open Graph title: "neatlogs — feedback to fix fast"
Why it matters
The title is the first line of every search result and the label on every tab. A brand-only title wastes the one piece of copy that reaches people who have not arrived yet.
Fix
Set the title to the brand plus the category, for example "Neatlogs, collaborative debugging for AI agents". Add the canonical link that is currently missing at the same time.
Body text under 14px
code · messaging clarity
mediumobservedquick win
13 block(s) of body copy set below 14px.
Evidence
13.5px — "We gave our external design agency access to one dashboard, but they w"
12.5px — "Created by @neatlogs on May 1. AI agent alerts and incident triage lan"
13px — "Quality regression in Revenue Q&A, trending higher than usual."
12.5px — "Distinguish paid members from guest collaborators"
9px — "Find anything about your traces"
9px — "Compare performance across versions"
9px — "List traces from user@example.com"
10px — "You are a billing support agent."
Why it matters
Small body copy pushes readers to zoom or give up, which costs time on page and comprehension.
Fix
Set body copy to 16px minimum, and no supporting copy below 14px.
Heavy pages lose visitors before they render, and the loss is worst on the mobile connections most visitors use.
Fix
Compress and correctly size images, serve modern formats, and defer non-critical scripts.
Visual and behavioral
10 issue(s) · scores 81/100
The imagery is decoration, not explanation
hero and closing section, desktop and mobile · art direction
highobservedmedium
Two illustrations carry the most valuable space on the page. The hero is a watercolour of the Golden Gate Bridge and a hillside. The closing section is a group of people flying through the air behind a laptop. Neither depicts an agent, a trace, a thread, a failure, or a fix. They have the smooth, generic look of assets generated to fill a slot rather than made to explain a product, and they would sit equally well on a site selling anything else.
Evidence
Hero illustration: Golden Gate Bridge and hillside, occupying the lower half of the desktop fold and roughly 40 percent of the mobile screen
Closing illustration: figures flying behind a laptop, above the only call to action in that section
Neither image contains a product surface, a trace, a thread, or any reference to agents
The actual product screenshots appear only from the second section onward
Why it matters
On a technical product the imagery is read as a signal of how well the team understands its own thing. Generic art in the two most valuable positions says the page was assembled rather than designed, and it spends the first impression on nothing while the strongest asset, the shared debugging thread, sits below the fold.
Fix
Replace both with owned imagery built from the product: the trace, the shared thread, the fix moving to Linear. Keep the hand-drawn warmth if it suits the brand, but every image should teach the visitor something they did not know a second earlier.
The headline is an outcome any developer tool could claim
hero, desktop and mobile · headline effectiveness
highobservedquick win
feedback to fix, fast is well written and well set, but it carries no information specific to this product. Swap in any bug tracker, error monitor, or code review tool and the line still works. The italic on fast is the only element doing differentiating work, and it is differentiating on speed, which every competitor also claims.
Evidence
Headline: "feedback to fix, fast"
Same page, section two: "See what happened, together"
Same page, section three: "See the issue. Ship the next step."
Why it matters
The headline is the single most read element on the site. Spending it on a claim competitors can copy word for word means the most valuable position on the page is doing no competitive work.
Fix
Trade the generic outcome for the specific one already written elsewhere on the site. "See what happened, together" is closer to the truth of the product than the current headline and is currently being spent on a section heading. Test it, or the domain expert and developer line, in the hero position.
The illustration takes the fold and carries no information
hero, worse on mobile · visual hierarchy
highobservedmedium
A watercolour of the Golden Gate Bridge and a hillside occupies roughly the lower half of the desktop fold and around forty percent of the mobile screen. It is well drawn and it has no relationship to agent debugging, traces, or teams. On mobile it pushes the two CTAs into the middle of the screen with the artwork below them, so the first thing the eye rests on after the buttons is a landscape.
Evidence
Desktop fold: illustration occupies the region below the CTAs to the fold line
Mobile 390x844: illustration occupies roughly the lower 40 percent of the first screen
Subject matter is a bridge and hillside, with no product or domain reference
Why it matters
The most valuable pixels on the site are being spent on decoration while the product screenshot, which is the strongest asset Neatlogs has, sits below the fold. Visitors who do not scroll never see the product at all.
Fix
Move the shared thread screenshot into the hero and demote the illustration to a background wash or drop it. The thread showing Sara, Marcus, and the Neatlogs agent working a real bug explains the category, the differentiator, and the product in one image.
The primary action does not win the fold
hero, desktop · cta prominence
mediumobservedquick win
Four button-styled controls sit above the fold. Get Started and Sign up share the same dark fill, so the hero's primary action and the navigation's primary action compete at identical weight. Book a Demo appears twice in a lighter outline style. No single control is clearly the one the page wants pressed.
Evidence
Four button-styled elements detected above the fold
"Get Started" and "Sign up" both use the same dark fill
"Book a Demo" rendered twice within the same screen
Why it matters
When two controls look equally primary, neither reads as primary, and click-through on the intended path falls.
Fix
Keep one dark filled button in the hero. Make the navigation Sign up a text link, and remove the duplicate Book a Demo from the navigation.
The page runs about twelve screens with pinned scroll sections
whole page, desktop · decision overload
mediumobservedmedium
Capturing the page took twelve full desktop viewports, and several sections pin in place while the content inside them advances, so the visitor scrolls without the page appearing to move. Two of those sections deliver a single sentence each. The scroll cost is high relative to the information delivered.
Evidence
12 viewport heights captured at 1440x900
Detect, Investigate, Act section pins while its three steps advance
See what happened, together section pins while the thread advances
Why it matters
Pinned scroll sections take control away from the reader. A visitor scanning for a specific answer, such as price or a comparison, cannot skim past them, and skimmers are the majority.
Fix
Keep the pinned treatment for the one section where the state change is the point, the Detect, Investigate, Act sequence, and let every other section scroll normally. Target roughly seven screens rather than twelve.
Both decision points are proof free
hero and closing CTA · social proof density
mediumobservedquick win
The two places a visitor decides, the hero and the closing start debugging together section, contain no evidence at all. There is no logo row under the hero buttons and nothing beside the final call to action except the free to start line.
Evidence
Hero: eyebrow, headline, subhead, two buttons, no proof element
Closing section: heading, one reassurance line, one button, no proof element
Why it matters
Proof placed away from the decision does not get used. Even strong evidence in the middle of the page will not reach the visitor at the moment they are deciding whether to click.
Fix
Put one line of proof directly beneath the hero buttons and repeat it beside the closing call to action. The same element in both places is fine.
Mobile pushes the product below four screens of scroll
mobile, 390x844 · responsive integrity
mediumobservedmedium
On mobile the hero copy wraps to four lines of eyebrow and headline plus four lines of subhead, stacking the two CTAs vertically and leaving the illustration to fill the rest of the screen. The first product screenshot does not appear until several screens down, and the trace and thread interfaces, which are dense two-column layouts on desktop, have to be scrolled horizontally or shrunk to fit.
Evidence
Mobile hero: headline wraps across two lines with the subhead running to four
CTAs stack vertically, Book a Demo above Get Started
20 tappable elements measured under the 44px minimum on a 390px viewport
Why it matters
A visitor arriving from a phone gets the least informative version of the page, and the product asset that does the persuading is furthest from them.
Fix
Cut the mobile subhead to one line, place the CTAs side by side, and lift a cropped detail of the thread screenshot into the mobile fold in place of the illustration.
Heading structure broken or multiple H1
code · visual hierarchy
mediumobservedquick win
2 H1 elements. Only one page-level heading should exist.
Evidence
feedback to fix, fast
neatlogs
Why it matters
Assistive tech and search crawlers use headings as the page outline. A broken outline makes the page harder to scan and to index.
Fix
Use exactly one H1 for the page topic and step heading levels down in order without skipping.
Touch targets under 44px on mobile
code · responsive integrity
mediumobservedquick win
20 tappable element(s) under the 44px minimum on a 390px viewport.
Evidence
"" — 117x32px
"" — 40x40px
"Book a Demo" — 117x40px
"Get Started" — 117x40px
"🎯 2" — 20x10px
"👀 1" — 20x10px
"" — 11x11px
"👍 1" — 20x10px
Why it matters
Targets under 44px cause mis-taps on mobile, which reads to the user as the site being broken.
Fix
Give every tappable element a minimum 44x44px hit area, using padding rather than font size.
Competing calls to action above the fold
code · decision overload
mediumobservedquick win
4 button-styled elements compete above the fold. One primary action should dominate.
2.51:1 (needs 4.5:1) — "Always use the latest pricing." lab(65.6464 1.53497 -5.42429) on rgb(250, 250, 250)
2.62:1 (needs 4.5:1) — "... see more" lab(65.6464 1.53497 -5.42429) on rgb(255, 255, 255)
Why it matters
Low contrast text is unreadable in sunlight, on older screens, and for the roughly one in twelve people with colour vision deficiency. It is also the single most litigated WCAG failure.
Fix
Raise text and background contrast to at least 4.5:1 for body copy and 3:1 for text 24px and above.
A visitor identification tracker runs on load, and is failing
network, all viewports · weak disclosure
mediumobservedmedium
The page calls a third party visitor identification service on load. During this scan that call returned a 429 rate limit error, so it is both running and currently broken. There is no cookie or consent notice presented before it fires. A privacy policy is linked in the footer, so this is a placement question rather than an absence of policy.
Evidence
429 response from https://opps-api.getwarmly.com/api/sessions/trpc/sessions.createSession
No consent or cookie notice shown before third party calls fire
5 distinct third party script domains loading
Why it matters
Identifying anonymous visitors before any consent is collected creates exposure under GDPR and similar regimes, and a European buyer evaluating a data-handling tool is exactly the visitor most likely to check.
Fix
Gate the identification call behind consent for European traffic, and confirm the privacy policy names this category of processing. Separately, fix or remove the failing call so it stops erroring on every page load.
A decorative asset is loaded from a third party demo host, and 404s
network, all viewports · misleading flows
lowobservedquick win
The page requests a noise texture from grainy-gradients.vercel.app, a public demo project on someone else's hosting, and that request returns 404. The visual it was meant to produce is silently absent, and the site has taken a hard dependency on a third party it does not control.
Evidence
404 response from https://grainy-gradients.vercel.app/noise.svg
2 JavaScript errors on load, both from failed resource requests
Why it matters
Console errors on load are the first thing a technically minded buyer notices when they open dev tools, which for an agent-debugging product is a likely thing for them to do. The dependency also means a stranger's project can change what your page renders.
Fix
Self-host the texture and remove the external reference. Add error monitoring so failed resources surface before a visitor finds them.
Missing title, description, canonical, or Open Graph
code · transparency
lowobservedquick win
No canonical link.
Why it matters
Missing metadata weakens search results and makes shared links render as bare URLs, costing click-through.
Fix
Add a unique title and meta description, a canonical link, and complete Open Graph tags including an image.
About these fixes
Every fix above is deliberately high level
Each one names what to change and why, which is enough to act on. What it leaves out is the reasoning underneath: which to do first given how people actually move through your funnel, what the replacement should say in your voice, and whether the change still holds once someone is inside the product rather than looking at the page.
Every recommendation we make is backed by a framework and a method, and tested against two things at once: whether it helps the person using the product, and whether it helps the business. A change that wins one and costs the other is not a fix, it is a trade nobody agreed to.
That is why we describe this as product design rather than visual work. If the thinking in this report is useful, the same thinking applied with your data in the room is where it compounds.
The product screenshots tell a real story, not a feature tour
The shared thread shows Sara spotting that add_member was used instead of the guest invite flow, Marcus asking the agent why, and the Neatlogs agent answering that the tool descriptions explain what the tools do but not when to use one instead of the other. That is a specific, credible failure with a specific cause and a proposed fix attached.
Keep it because: Almost every competitor shows a dashboard. This shows a team reasoning through a bug together, which is the actual product. It is the most persuasive thing on the site and it should be doing more work, higher up.
Simple enough for domain experts. Detailed enough for developers.
This single line names the wedge: a debugging surface that a non-engineer can read and comment on without the engineer losing the trace-level detail they need.
Keep it because: This is the sentence the whole site should be organised around. It answers who else can use this, which is the question that separates Neatlogs from every trace viewer built only for engineers.
The FAQ answers the hardest objection head on
Question two asks how Neatlogs is different from observability tools and answers it cleanly: standard observability tells you what happened, Neatlogs is built around the debugging workflow with shared traces, comments, evaluations, tasks, and suggested fixes.
Keep it because: Most teams dodge the comparison question. This one answers it in two sentences, and the answer is good. It is in the wrong place, but the thinking behind it is right.
Detect, Investigate, Act is a clean three-step spine
Catch the spike early, find out what went wrong, move from insight to action. Three steps, one sentence each, each paired with a matching state in the product panel beside it.
Keep it because: It gives a reader a mental model of the product in about eight seconds. That structure should survive any redesign even if the surrounding page changes completely.
The page is genuinely fast
First contentful paint at 408ms and DOM ready at 262ms, on a page carrying 80 images and a scroll-driven narrative. That is a well-built front end.
Keep it because: Speed is the one trust signal you cannot fake, and for a developer audience evaluating an infrastructure tool it is read as competence.
Legal and support routes are all present
Privacy policy, terms of service, a glossary, and a Contact Support route are all linked in the footer, and the still need help section offers a call.
Keep it because: This is the baseline most early stage sites miss, and missing it costs trust with exactly the buyers who read footers.
Section 04
Do these three first
01
Move the category and the differentiator into the hero
A first-time visitor learns what the product is and why it is not just observability within the first five seconds instead of fourteen screens down. This is the largest single drop-off on the page and the copy already exists, in the FAQ, so it costs a paragraph rather than a redesign.
02
Fix the closing call to action so it matches its own promise
Visitors who read free to start, no credit card required and reached for it can actually start. Recovers the highest intent traffic on the page and removes a contradiction that currently reads as bait.
03
Put the product screenshot and one line of proof above the fold
Replaces a decorative illustration with the strongest asset on the site, and gives a buyer the evidence they need at the moment they decide. Also fixes the mobile experience, where the product is currently several screens away.
This is where it starts, not where it ends
Before you act on these
These three are the order, not the plan
Sequencing three changes is straightforward. Knowing which one to spend a sprint on, what the replacement copy should actually say, and how to tell within a fortnight whether it worked is a different question, and it depends on numbers this report cannot see.
That is the part we would want a conversation for. Thirty minutes, your funnel data on screen, no pitch.
Static capture only. Anything behind a click, a login, or a payment step was not exercised and is flagged
as a risk rather than a confirmed fault. Findings marked "likely" or "possible" need a live walkthrough
to confirm. Scores are the framework's weighting of severity and confidence, not an industry benchmark.
Generated by UX Signal Studio using the UX Risk framework.
Section 07
Art direction
Design verdict
Not optimal
Opinion, not measurement
The hero is a watercolour of the Golden Gate Bridge and a hillside. The closing section is an illustration of people flying through the air behind a laptop. Both are well drawn. Neither shows an agent, a trace, a team, or anything else the product touches.
This is where we part company with the rest of the site, and it is a view rather than a measurement. The illustration is decoration standing in the place where the product should be. It reads as art commissioned to fill a space, not art made to explain something, and on a technical product that gap registers as a company that has not decided what it is. You already own the better image: the shared thread where Sara, Marcus, and the Neatlogs agent work out why add_member was chosen over invite_guest. That is the product, it is specific, and it is sitting below the fold behind a bridge.
What we would do instead
Retire the stock-feeling landscape and build a small, owned visual language from the product itself: the trace, the thread, the fix moving to Linear. Same warmth and hand-drawn quality, but every image carrying information. The cover shows what the product does, the footer shows what it produces, and nothing on the page is there only to fill a gap.
Everything else in this report is measured or quoted. This section is not. It is a design judgement,
offered because the gap between the imagery and the product is the kind of thing a scan will never
flag and a buyer will always feel.
Appendix
How the score is built
Each check starts at 100 and is reduced by its most severe finding. A category is the average of
its checks. The overall score weights the three categories at 40, 35, and 25 percent, then takes
five points for every unresolved critical finding, because a critical is a business consequence
that strong scores elsewhere should not wash out.
This report is a combination of automated and human work. The capture, the code analysis, and the legal and compliance probes are measured directly from the page. The judgement calls, what counts as a finding, how severe it is, and what to do about it, are ours. AI is used for the measurement pass and to help draft, which is what makes it possible to read every screen rather than skim. That also means some results can be wrong. Treat anything here as checkable rather than final, and tell us if a finding does not hold, because we would rather correct it than defend it.
Section 09
This is the surface
Your whole site next, not just this page
A scan reads one page the way a stranger does. It can tell you the category is missing from the
hero and that the closing call to action contradicts its own copy.
What it cannot do is decide what the product should promise, or design the flow that keeps that promise once someone is inside it. Everything in this report is above the signup. The compounding return is below it.
What the engagement adds
Every screen, not one. The full site, the onboarding flow, and the paths this scan cannot reach because they sit behind a signup.
Designed, not just diagnosed. We do the work, so the fix ships rather than joining a backlog as a note.
Your analysts in the room. We read the funnel data with the people who own it, so we are arguing from numbers rather than instinct.
Experiments, not opinions. Each change ships as something measurable, so you know within a fortnight whether it worked.
What a scan reaches
One public page, as a first-time visitor sees it
Copy, hierarchy, and the first five seconds
Accessibility, performance, and consent in the markup
