Confidential — Internal Audit

Klaut.id — Pre-Launch Website Audit

Engagement classification: Brand-and-Copy diligence, scored to Big-Four web-audit methodology
Deloitte-grade methodology
Audited URL: klaut.id
Issue date: 14 May 2026
Lead reviewer: Claude (acting auditor)
C+
59 / 100
Conditional Recommendation

Strong foundation, polish required before enterprise outreach.

The brand identity, video hero, and four-service architecture present competently. Underneath, the site carries identifiable AI-writing patterns, factual inconsistencies that an enterprise procurement reader will notice within seconds, and a single-page architecture that will not absorb the case-study and career content the business plan calls for. Specific issues are itemized in the Findings tab; remediation is sequenced in the Verdict tab.

17 Issues identified
2 Critical · 5 High
72% AI-text fingerprint
Typo / Grammar
62
9 issues · 1 critical
Copy & Wording
61
Buzzword-dense, register inconsistent
Visual Design
78
Strongest dimension
Message Clarity
58
Positioning blurred by jargon
AI-Writing Fingerprint
35
Lower = more AI-detectable
Scalability
55
Single-page wall ahead
Conversion Path
64
CTAs work, route to email
Recommend to Buyer?
58
Conditional

Seven-dimension report card

Top three risks

  1. 1
    Email-domain mismatch
    Footer and mailto links route to hello@klaut.ai while the live domain is klaut.id. If .ai is not owned, every inbound enquiry bounces.
  2. 2
    Unbacked claim wording
    "5X median time to production" reads as 5x slower, not faster. "60% operational cost saved" prefixed with "Save" is redundant. Enterprise readers parse these.
  3. 3
    AI-text fingerprint at hero
    Phrases like "vertically-integrated transformation ecosystem" and "battle-tested in regulated environments" trip every detector and signal generated copy to a sophisticated buyer.

Severity distribution 17 findings classified

Audit philosophy

This audit is scored against the standards of a sophisticated enterprise reader — a CIO, a procurement lead, a CFO's chief of staff. That reader spends 5–8 seconds on a hero before deciding whether to scroll, and 30–45 seconds on the first scroll before deciding whether to book a call.

A "C+" grade does not mean the site is poor work. It means the site is competent on craft (visuals, structure, hierarchy) but loses points on the dimensions that matter most for closing a 5–6 figure consulting engagement: factual precision, register consistency, and proof.

The good news is the gap is closable in roughly 6–10 hours of focused copy and asset work, without changing the design system or build. Sequencing is in the Verdict tab.

Typo & grammar findings — 9 items

Click any row to expand the original text and the suggested correction.

Location Issue Severity

Register inconsistency map

The site oscillates between three voices: enterprise-formal, casual-friendly, and pop-culture-jokey. This is the single largest credibility leak. A CIO reading the homepage cannot tell what kind of firm Klaut is.

Voice samples

Same site, three different writers.

FORMAL
"Klaut is a single, vertically-integrated transformation ecosystem."
CASUAL
"You need customer service? we got it. You only need CRM? no problem."
JOKEY
"Code audit to make sure your AI agent will not become a terminator in the future."

Banned-word density

Words flagged by the firm's internal anti-AI-writing rulebook that nevertheless appear on the site. Each one weakens the perception that a human wrote the page.

Notable wording problems

Phrase as writtenProblemSuggested rewrite
"AI Ecosystem Designed for you, personally" "Personally" contradicts "Enterprise Grade" — softens the positioning rather than reinforcing it. "AI Ecosystem — Built around your operation"
"Battle-tested in regulated environments" Klaut was founded in 2025. Unsubstantiated claim. Enterprise buyers will ask which regulators, which environments. Either cite specifics ("deployed in [X] regulated banking workflow") or remove.
"AI leverage points" "Leverage" as a verb/noun is on the internal banned list. Buzzword filler. "Where AI delivers the most operational lift" or simply "where AI helps most"
"Deploy your message" CTA on a mailto link. Cute but the verb mismatches the action. Reader hesitates. "Send a brief" or "Start a conversation"
"From a cup of coffee to autonomous system" Tonally pleasant but missing an article and ungrammatical. Sells low-stakes when the section is the six-phase enterprise process. "From a first conversation to an autonomous system"

Design dimension scores

Design summary

Design is the strongest dimension of the site. The hero video, monochrome wordmark, scrolling capability marquee, and numbered service blocks read as a 2026-era enterprise SaaS marketing site. The K-mark and wordmark logo system is competent and consistent.

Three concerns prevent a higher score. First, footer routing: Career → #cao, Privacy → #, Terms → #, Security → # — these are placeholder routes. A diligence-minded buyer who clicks any of them sees the page jump to nowhere. Second, the case-study "View work →" links do not appear to route to actual case studies, undermining the "7 SHIPPED" and "4 SHIPPED" proof points. Third, the integration node-graph section contains a visible string "1. 00Hover or click a node" which looks like an unrendered template variable.

None of these are aesthetic failures. They are completion failures, which is worse on a launched site.

Design issue log

ElementIssueSeverity
Footer · Career linkRoutes to #cao instead of a careers pageHigh
Footer · Privacy / Terms / SecurityAll three resolve to #, no policies linkedHigh
Industries · View work CTAsSix "View work →" CTAs with no destination, despite "7 SHIPPED" / "4 SHIPPED" claimsCritical
Integration graphicVisible string "1. 00Hover or click a node" appears unrenderedMedium
Nav · Book a callSays "Book a call" but routes to mailto, not a scheduler. Sets false expectation.Medium
Service tags · CAO & GovernanceBoth list "Enterprise UMKM Personal" — UMKM is Indonesian (SME). Inconsistent with English-language page.Medium

AI-writing fingerprint score

Lower is better. A score below 50 means a competent reader will assume the page was written by an LLM with light human editing.

Five-second test

Verdict: Partially passes.

A first-time visitor learns within five seconds that Klaut is an enterprise AI company in Indonesia with a video-led brand. What the visitor does not learn:

  • That Klaut does not build AI — it integrates and manages vendor stacks
  • The general-contractor analogy that distinguishes Klaut from a model-builder
  • The three revenue streams (Discovery → Implementation → Retainer)
  • Why a mid-market enterprise should choose Klaut over Accenture or a freelance AI consultancy

The positioning is blurred by buzzwords. The general-contractor framing in the founder's playbook is the strongest differentiator and is currently not on the site.

AI-writing tells detected

PatternExample from the siteWhy it reads AI
Buzzword stacking "single, vertically-integrated transformation ecosystem" Four abstract nouns chained. No human writes this; pattern-matching does.
Three-part parallel cadence "One contract. One roadmap. One accountable team." LLM-favorite rhythm. Acceptable once; the site repeats it (Consult/Construct/Deploy, three Outcomes triplet).
Vague-uplift sentences "Harmonizing Human & AI relationship for a better future" Mission-statement filler. Says nothing specific.
Identical service-card cadence Each of the six industry blocks starts with a verb + buzzword stack of the same length Reads as a templated prompt response.
Unsubstantiated absolutes "battle-tested in regulated environments" Claim without referent — typical LLM-generated authority signal.

Counterpoint: the CEO quote ("Most people think AI will replace their job and business. In fact, it will help you to do better than ever.") reads human — slightly awkward, conviction-led. That voice should be the model for the rest of the body copy.

Architecture readiness

The single-page wall

Klaut.id is currently a single-page anchor-link site. Every navigation item — About, Case Studies, Why Klaut, Book a call — scrolls to a section of the same page. This is a competent launch pattern but it has a hard ceiling.

The business plan explicitly calls for case studies, vertical-specific playbooks, hiring pages, partnership pages, and a media presence. None of these are servable from a single page without becoming a scroll-fatigue document. SEO compounds the problem: one URL, one title, one meta description has to win for every keyword the firm wants to rank for.

The good news: nothing in the current design system blocks expansion. The hero, marquee, and service blocks become the home page; everything else becomes routed pages.

Scalability dimensions

DimensionCurrent stateScales to 10x traffic?Scales to 10x content?
Hosting / deliveryStatic siteYesYes
Information architectureSingle-page anchorsYesNo
SEO surface area1 URL, 1 title tagLimitedNo
Case-study capacity"View work →" links lead nowhereNo
Hiring / careersCareer link routes to #caoNo
Legal / compliancePrivacy / Terms / Security stubsNo
Multilingual (Bahasa)English-only with one UMKM mentionNo
Content managementHand-coded (inferred)YesPainful

Recommendation: Conditional Yes

I would recommend klaut.id to an enterprise buyer after the seven Priority-One fixes below are landed. As shipped today, the site is presentable for warm-introduction conversations where the buyer already trusts the founder, but is not ready for cold outbound or for a procurement reader who scrolls without context.

Remediation roadmap Sequenced by impact-to-effort ratio

Priority One — Fix this week (≈4 hours)

  1. 1
    Reconcile the email domain
    Either own klaut.ai and forward, or change every mailto to @klaut.id. Right now inbound enquiries may bounce.
  2. 2
    Rewrite the broken stat cards
    "5X median time to production" → "5x faster to production". "Save 60% operational cost saved" → "60% reduction in operational cost".
  3. 3
    Strip the AI-fingerprint phrases from the hero and About blocks
    Replace "vertically-integrated transformation ecosystem" and "battle-tested in regulated environments" with concrete operator language.
  4. 4
    Remove the "terminator" line
    Tonally incompatible with an enterprise compliance buyer. Replace with "Independent code review and audit certification for every deployment."
  5. 5
    Fix the broken UI string in the integration graphic
    "1. 00Hover or click a node" should not be visible.
  6. 6
    Add stubbed Privacy / Terms / Security pages
    Even one-paragraph placeholders. The empty # hrefs are visible diligence flags.
  7. 7
    Standardize the casual sections
    "We got it." → "We can deliver it." / "no problem" → "No problem." Capitalize sentence starts; one register throughout.

Priority Two — Land within 30 days

  1. 1
    Build two real case-study pages
    One Financial Services, one Marketplace. Activate the "View work →" links. Even sanitized, anonymized writeups outperform "7 SHIPPED" with no detail.
  2. 2
    Surface the general-contractor positioning
    Add a one-section explainer: "We don't build AI. We coordinate the AI vendors that do — and stand behind the result." This is your differentiator. It is currently invisible.
  3. 3
    Decide on Bahasa
    Either commit to English-only and drop "UMKM", or build a Bahasa toggle. Mid-market Indonesian buyers will respond to native-language pages.
  4. 4
    Wire a real scheduler to "Book a call"
    Cal.com or Calendly. The mailto bait-and-switch costs you the lead.

Priority Three — Architectural (60–90 days)

  1. 1
    Migrate to a routed multi-page architecture
    Home, Services (4 sub-pages), Case Studies (index + details), About, Careers, Contact. Keeps the home-page visual system, unlocks the SEO surface area.
  2. 2
    Add a Resources or Insights section
    Two short essays from the founder, published quarterly. Establishes authority faster than any case study and supplies referral context.
  3. 3
    Instrument analytics
    Hero-video play rate, scroll depth, CTA click-through, form completion. Without this, every copy change is faith-based.

Final score tally

Typo & Grammar
62
Copy & Wording
61
Visual Design
78
Message Clarity
58
AI-Writing Resistance
35
Scalability
55
Buyer-Recommendation Readiness
58
Overall composite (weighted)
59 / 100 · C+