Loading Report View…
0/ 100
gnani.ai Website Audit Report
gnani.ai scored 79 out of 100 in this SavageAudit website audit covering Performance, SEO, Design, Copy, User Experience, and Conversion. The report uses captured website evidence to explain what is working, what is limiting the experience, and which improvements deserve attention first.
“Enterprise voice AI, wearing a tuxedo over a power outage.”
Gnani AI looks like it should be auditioning for an enterprise keynote, but the homepage keeps tripping over its own feet. The message is strong; the execution keeps wandering into the hallway and forgetting why it was invited.
https://www.gnani.ai/
TECHNICAL SUMMARY
The site’s strongest problem is speed: LCP is 12.2s, interactive lands at 13.4s, and the page weighs 9,222 KiB, which is an outrageous suitcase for a homepage selling enterprise voice AI. SEO and social metadata are largely solid, but the real-world experience is undermined by 2 console errors, 12 broken assets, 27 missing alt texts, and 4 forms with 13 inputs—so the fix-first move is obvious: cut the heavyweight assets and repair the broken scripts before those demo clicks evaporate into patience-loss and revenue drag.
Where It Burns
6 categoriesWebsite Audit Findings for gnani.ai
SavageAudit identified 37 visible findings and recommendations across 6 audit categories. Each section below combines the category score with observed evidence and prioritized fixes.
5101 PERFORMANCELCP 12.2s9502 SEOSEO Score 1009503 DESIGNImages 1029104 COPYWord Count 1,5267005 UXJS Errors 29006 CONVERSIONCTA Count 5
01
PERFORMANCE AUDIT
51
LCP12.2sInteractive13.4sTotal Byte Weight9,222 KiBTBT430 ms
- 🔥Your LCP of 12.2s is not a hero metric, it’s a hostage situation. In voice AI, that’s longer than the market’s patience and several support calls.
💀
A heavyweight homepage is trying to sell speed while dragging a suitcase full of bricks.
- →Compress the hero assets and convert them to WebP/AVIF so the homepage stops arriving like a delayed weather front.LCP: 12.2 sImpact: highEffort: medium
Why this matters
The LCP is 12.2s, so the hero is clearly too heavy and delaying the first meaningful render.
- →Defer non-critical JavaScript and trim render-blocking resources so the main thread can breathe before the demo pitch starts.TBT: 430 msImpact: highEffort: medium
Why this matters
430ms of TBT shows JavaScript is still chewing up responsiveness when it should be getting out of the way.
- →Turn on CDN caching with long TTLs for static assets so first-load latency stops acting like a bureaucratic queue.FCP: 4.7 sImpact: mediumEffort: low
Why this matters
FCP is 4.7s, and the page is carrying enough bytes that better caching can materially reduce the wait.
02
SEO AUDIT
95
SEO Score100Schema TypesOrganization, Person, ContactPointMissing Alt Text27AI CrawlersMixed 1/4
- 🔥The SEO setup is mostly fine, which is almost suspiciously competent for a homepage carrying 27 missing alt texts. Even the search bots are being handed a slightly lazy tour.
💀
SEO is doing the job, but the metadata wardrobe is missing a few very enterprise-looking pieces.
- →Add descriptive alt text to every meaningful image so accessibility and search indexing stop taking the scenic route.Missing Alt: 27Impact: mediumEffort: medium
Why this matters
27 missing alt texts is too many for an enterprise homepage and weakens accessibility plus image understanding.
- →Review robots.txt so GPTBot, ClaudeBot, and Google-Extended can crawl the pages you actually want cited.AI Crawlers: Mixed 1/4Impact: mediumEffort: low
Why this matters
With only 1 of 4 AI crawlers effectively allowed, discoverability in AI answers is being artificially limited.
03
DESIGN AUDIT
95
Images102Missing Alt27H1 Count1Font Families3
- 🔥102 images and 27 missing alt texts is a lot of visual content to be this casually under-described. The layout looks intentional, but the accessibility notes read like someone left half the labels in the truck.
💀
Clean styling, but the design still leaves too much work to the content and too little to the screen reader.
- →Add supporting imagery where the layout feels sparse so the page doesn’t look like it borrowed confidence from the whitespace.Images: 102Impact: mediumEffort: medium
Why this matters
With 102 images already on the page, the visual system should be doing more to support key sections instead of leaving gaps in the story.
- →Fix missing alt text across the image set so accessibility audits stop failing on avoidable basics.Missing Alt: 27Impact: mediumEffort: medium
Why this matters
27 missing alt texts is a direct accessibility and semantics problem, not a style preference.
- →Test spacing and alignment across breakpoints to make sure the already-clean layout stays clean when the viewport changes its mind.Viewport: PresentImpact: lowEffort: medium
Why this matters
Viewport support is present, but responsive behavior still needs verification to protect the design across devices.
04
COPY AUDIT
91
Word Count1,526Sentence Count126H1 Count1CTA Count5
- 🔥1,526 words and 126 sentences on a homepage is not a message; it’s a memoir with a CTA at the end. Enterprise buyers don’t need a novel before they can book a demo.
💀
The writing knows what it wants to say, then spends way too long saying it.
- →Tighten the headline so it states the value proposition in one sharp sentence instead of a stitched-together label.H1 Count: 1Impact: mediumEffort: low
Why this matters
The single H1 should do the heavy lifting, and right now it’s carrying the brand line instead of the value.
- →Break dense sections into shorter sentences and bullets so enterprise readers can scan instead of excavating.Avg Words/Sentence: 12Impact: lowEffort: medium
Why this matters
The page has 126 sentences and 1,526 words, which is a lot of reading for a homepage that needs faster comprehension.
- →Make the primary CTA stand out and reduce repetitive CTA wording so the page stops sounding like it’s echoing itself.CTA Count: 5Impact: mediumEffort: low
Why this matters
Five CTA instances with only two labels makes the call-to-action hierarchy feel diluted rather than decisive.
05
UX AUDITUSER EXPERIENCE AUDIT
70
JS Errors2Broken Assets12Long Tasks7CLS0.022
- 🔥Two console errors, including "Cannot read properties of null (reading 'style')," is the page admitting it has a small internal crisis in public.
💀
The experience mostly works until the page starts breaking things and apologizing in console logs.
- →Fix client-side JavaScript errors so interactions stop failing silently in production.JS Errors: 2Impact: highEffort: medium
Why this matters
Two console errors, including a null-style read, are user-facing instability waiting to happen.
- →Repair failed asset URLs and deployment paths so scripts, images, and stylesheets load cleanly.Broken Assets: 12Impact: highEffort: medium
Why this matters
Twelve broken assets is enough to create visible breakage and undermine confidence in the page flow.
- →Reduce the number of form fields to only the essentials so the page stops asking enterprise buyers to fill out a tax return for a demo.Form Fields: 13Impact: mediumEffort: medium
Why this matters
Thirteen fields is a lot of effort for a book-a-demo journey and raises unnecessary completion risk.
06
CONVERSION AUDIT
90
CTA Count5Unique CTA Labels2Forms4Social Proof4
- 🔥There are 5 CTA instances, but only 2 unique labels, so the page is basically chanting "Book a Demo" and "Start Building" until one of them works. That’s not persuasion, that’s repetition with branding.
💀
The page wants a demo, but it keeps making the ask feel like a second job.
- →Place a primary CTA above the fold that matches the core offer so visitors don’t have to decode the page first.CTA Count: 5Impact: mediumEffort: low
Why this matters
Five CTA instances with only two labels suggests the primary action needs clearer hierarchy, not more duplication.
- →Keep a low-friction contact path visible for buyers who are not ready to book immediately.Contact Path: PresentImpact: lowEffort: medium
Why this matters
The contact path exists, but it should remain easier to reach for enterprise buyers who want a softer next step.
- →Move social proof closer to the CTA so testimonials, logos, or stats can do the convincing before the form does the asking.Social Proof: 4Impact: mediumEffort: low
Why this matters
Four social proof items are available, but their placement needs to work harder at the decision point.
Audit overview
gnani.ai Audit Summary
The site’s strongest problem is speed: LCP is 12.2s, interactive lands at 13.4s, and the page weighs 9,222 KiB, which is an outrageous suitcase for a homepage selling enterprise voice AI. SEO and social metadata are largely solid, but the real-world experience is undermined by 2 console errors, 12 broken assets, 27 missing alt texts, and 4 forms with 13 inputs—so the fix-first move is obvious: cut the heavyweight assets and repair the broken scripts before those demo clicks evaporate into patience-loss and revenue drag.
Based on the category scores in this report, SEO is the strongest area while Performance has the most room for improvement.
How SavageAudit Evaluated gnani.ai
SavageAudit evaluated the public website at https://www.gnani.ai/ across performance, technical SEO, design, copy, user experience, and conversion signals. Scores and recommendations reflect the website evidence available when this audit was generated on July 4, 2026.
Similar Website Audit Reports
Compare this result with other public SavageAudit reports that received a similar overall website score.