01 Executive Brief
Executive Summary & Risk Map
Two pages for stakeholders: the core risks, what each one is costing, and the recovery timeline.
SEO Audit Services that go past the checklist
Broken links and missing tags rarely explain why traffic fell. This SEO audit service finds the issues underneath: discounted links, pages with nothing to add, templates Googlebot never renders, a category with no hub page.
- 7 modules, which of these is actually hurting us
- Problem → Cause → Impact → Fix, how every finding is written up
- 0 generic recommendations, every finding names a URL you can open
- Technical & Rendering Audit
- Backlink & Link Equity Audit
- Content & E-E-A-T Audit
- Entity & Schema Audit
- AI Visibility Tracking
- Powered by Algorithm
What the SEO audit services include, seven modules
Core Capabilities
An SEO audit is a one-off review of what blocks a site in organic search: technical access, on-page content and the backlink profile. Ours runs on Algorithm, our own platform, fed by Screaming Frog, Search Console and server logs, and each finding names a report, a query set or a URL you can open.
Technical SEO Audit
Crawl, JS rendering, indexation, Core Web Vitals
Backlink Audit
Discounted or toxic links, with the evidence per domain
On-Page & Content Audit
Intent match, cannibalization, content vs the top 10
Entity & Schema Audit
Category hubs, schema per URL, Knowledge Graph entry
Update Loss Forensics
Which update, which section, which pattern
SXO & Intent Satisfaction
Post-click UX: pogo-sticking & conversion paths
AI Visibility (GEO)
Brand citations in AI Overviews, ChatGPT & Perplexity
Symptom → Diagnosis
Seven symptoms the SEO audit traces to a cause
The symptom as you'd describe it, the cause we usually find behind it, and what we check to confirm it.
01
Discounted Backlinks
Symptom
Traffic drops with zero manual penalties in Search Console.
+
01
The Real Problem
Part of the link profile stopped counting. Since the 2022 link spam update Google neutralises paid and network links algorithmically, and Search Console sends no notice when it does.
What We Check
Which referring domains carry the signals Google discounts, paid-post hubs and shared network fingerprints first, and the month their equity stopped arriving.
02
Content With Nothing to Add
Symptom
Expert content steadily loses visibility after Core Updates.
+
02
The Real Problem
The pages say what the top 10 already says. Google's helpful content guidance asks whether a page adds something beyond existing sources; a well-written rewrite of the consensus doesn't.
What We Check
Your page against the top 10 for its query: which facts, examples and sources are yours alone, and which sections repeat the pages already ranking.
03
Category-Wide Decline
Symptom
Rankings decline across whole categories over several update cycles, with no single page to blame.
+
03
The Real Problem
No page owns the category. The topic is spread across posts, the hub is thin or missing, and schema names the business differently from template to template. Google reads a category from its hub and the links into it.
What We Check
Hub page per category and the internal links into it, Organization and Product/Service schema across templates, and whether the brand has a Knowledge Graph entry at all.
04
JS Rendering Gaps
Symptom
Green Core Web Vitals, but Googlebot fails to index deep pages.
+
04
The Real Problem
Links and content that exist only after JavaScript runs. Googlebot renders in a second pass with a budget, and templates deep in the site don't make it to that pass.
What We Check
Rendered DOM against raw HTML per template, the links that exist only after rendering, and which templates Googlebot actually requests in the server logs.
05
Keyword Cannibalization
Symptom
New pages push legacy, high-converting pages down instead of growing traffic.
+
05
The Real Problem
Two pages answer one query, so Google alternates between them and neither holds the position. Usually a new post written for a query a service page already owned.
What We Check
Query-level history from Search Console: URLs swapping on one query, page pair by page pair, and which of the two should keep it.
06
Missing From AI Answers
Symptom
Rankings look stable, but high-intent traffic drops as engines stop citing your brand in AI answers.
+
06
The Real Problem
AI engines cite passages they can attribute: a named source, a number, a first-hand example. Pages written as marketing copy give them nothing to quote, so a competitor's page is cited instead.
What We Check
Brand citations per query across ChatGPT, Perplexity, Gemini, Claude, Grok and Google AI Mode, and what the cited pages have that yours lack.
07
Thin Content Drag
Symptom
Flagship pages keep improving, yet the whole domain sinks. The site seems capped regardless of page quality.
+
07
The Real Problem
A large share of the site is thin, duplicated or outdated. Google's helpful content classifier was site-wide, and since March 2024 it runs inside core ranking. Five improved flagship pages don't offset two hundred weak ones.
What We Check
Share of thin and duplicate URLs per section, index bloat (URLs indexed vs URLs that earned a click in 12 months), and which sections to prune, merge or noindex.
Platform Access
How the SEO audit services run: five reports, on your own login
The audit runs on Algorithm, our own platform, with Screaming Frog, Search Console and server logs feeding it. The report comes as PDF, Excel and Word, and your team gets a platform login for the audit and after it, so any number in the report can be opened at source.
Link Equity & Toxicity
DR inflates easily. The backlink audit isolates the authority a link actually carries, with the evidence behind each flag.
What you see
- Which links work, and which only look the part. Referring domains sorted into two lists: the ones that carry authority, and the ones carrying signals Google discounts, paid-post hubs and off-niche filler among them. Evidence per flag.
- Where your anchor profile stopped looking natural. Anchor distribution against the over-optimization line, money-keyword spikes by date, and a disavow file reasoned per domain.
- Your link network, mapped. Domains sharing hosting, ad or analytics IDs, or ownership, one operator voting for you many times, named with the shared fingerprint.
Shared ad and analytics IDs are listed per domain, so a seller can't dispute the flag.
Semantic Site Map & Page Intent
Pages are grouped by the queries they share and the sections they duplicate: what competes with what, what should link to what, and whether each page is the type Google ranks for its query.
What you see
- Pages competing for one query, down to the duplicated section. Two articles, or an article and a service page, that Google can't choose between. Paired, with the call: merge, differentiate or link.
- Core pages starved of internal links, with their donors. The pages at the centre of your topics sitting at the bottom of internal authority, each with the strong pages that should link to them, and the anchor.
- Wrong page type for the query. If the top is guides and yours is a landing page, no on-page work fixes it. Format and intent mismatches listed with the page type that wins.
Most “content problems” turn out to be structure problems: the wrong page, or two pages, on one query.
Content Authenticity & E-E-A-T
AI search engines attribute facts. The content audit measures the citability, first-hand experience and proof signals that word count misses.
What you see
- Your citability against the pages that outrank you. Independent ideas, self-contained passages, named sources, attributed quotes, counted on your page and on theirs. These are what AI answers lift.
- Each claim on your service pages matched to its proof, or flagged. “Experts”, “15 years”, “trusted by”, paired with the passage that backs it, or listed as unsupported. Plus the anonymous team blocks.
- AI-written share and staleness, page by page. Which pages read machine-written, which haven't been touched in years, which lack a dated author, ranked by fix order.
The same check runs on the pages that outrank you, so the bar is visible next to your score.
Full Coverage
Before any diagnosis, the whole domain on one screen: organic traffic and positions, competitors, backlinks and history, pulled fresh when you ask, with the raw data under each number.
What you see
- Where your rankings sit, and where they're moving. Keywords split by position band (top 3, 4–10, 11–20, beyond) with monthly history and the new / up / down / lost movement; intent and SERP features per query.
- Who is taking your organic traffic. Competitor analysis with the generic giants filtered out: rivals ranked by keyword overlap, the keyword gap against your closest one, traffic share and which of them is growing.
- The signals other tools skip. Traffic concentration, reachable traffic at positions 4–20, Link Risk from your live backlink profile, share of queries with AI Overviews, and the domain's past life from the Wayback Archive.
The export sits under each table, one click away.
AI Visibility Diagnostics
AI engines quote what they can attribute. The audit tracks where you're cited across six engines, and whether your pages give them a reason to.
What you see
- Where you're cited across six AI engines, and who is when you're not. ChatGPT, Perplexity, Gemini, Claude, Grok and Google AI Mode, per query, with the competitor holding your slot named.
- Recommendations separated from luck. Each query run three to five times: 5/5 is a recommendation, 1/5 is noise. You act on the stable ones.
- The pages AI cites, checked against yours. The sources engines pull from on your queries, run through the same citability check as your page.
Each engine is asked in the buyer's phrasing: “best X for”, “X alternatives”, “X vs Y”.
More in the platform
Engineering tools for the scenarios where a mistake costs traffic: migrations, JS-heavy apps, post-update drops, due diligence.
Internal Link Analyzer
Orphan pages, click depth, broken links, redirects, noindex and canonical mismatches across up to 5,000 URLs. Content links and navigation links are counted separately, because a link in the body and a link in the footer don't carry the same weight.
Pre-Migration
Complex Web Apps
JS-Rendered Crawl & Technical Check
Headless-Chrome crawl for React, Next.js and Vue sites, so you see links and content as Googlebot renders them on mobile. Per URL: indexability, Core Web Vitals and page speed, structured data, hreflang, AI-crawler access.
Complex Web Apps
Redirect Chain & Loop Inspector
Bulk detection of multi-hop redirect chains, loops and canonical conflicts across the URL map, up to 500 URLs per run.
Pre-Migration
M&A Due Diligence
Indexation Coverage Check
Indexed or not, URL by URL across the sitemap, in Google and Bing. The first thing to read after a drop.
Post-Update Drops
Schema Validation & Generator
Structured data validated per URL, and valid JSON-LD generated for 15 types with required and recommended properties checked as you type.
Complex Web Apps
Pre-Migration
Content Audit & AI-Footprint Scan
Each article graded on depth, E-E-A-T signals and genre fit; AI-generated share per page; keyword cannibalization and freshness across the whole blog. Keep, rewrite or delete, as a named list.
M&A Due Diligence
Post-Update Drops
Deliverables
What you get, and your team can ship
Three things leave the audit with you: a two-page brief with the reasoning written out, an Excel engineering backlog as the prioritized action plan, and platform access to recheck the findings yourself.
02 Dev Tickets
Engineering Backlog
Tickets a developer can pick up as written, with the affected URLs, prioritised High / Med / Low by impact.
03 Real-Time Platform
Algorithm Platform Access
Live dashboard for indexing, entity status and rankings, yours during the audit and after handoff.
04 14-Day Async Support
14-Day Engineering Support
Two weeks of written engineering support after handoff, with direct Q&A with our Lead Architect while your team ships the backlog.
The Difference
How this differs from the audit you're used to
Regular Automated Audit
Neon SEO Audit
FocusSuperficial HTML notices & broken links
FocusCore Update drops & architectural flaws
ScopeTechnical, on-page and off-page as three separate checklists
ScopeThe same three, read together, because the drop usually sits where they overlap
ToolsRaw exports from third-party crawlers
ToolsGSC, Screaming Frog & logs, unified in our Algorithm platform
Output100-page PDF with unprioritized tasks
OutputEngineering backlog prioritised by impact, ticket by ticket
DeliveryAccount manager passing PDFs
Delivery14 days of async engineering support with Lead Architect
Workflow
Three weeks from kickoff to a clear picture of what's wrong
Fixed scope. Fixed three-week timeline. Engagements start at $1,500.
Week 1
Data Integration & Baseline
Search Console and GA4 connected, server logs pulled where you have them, a crawl of the whole site in Algorithm. This is the baseline everything gets compared against.
Week 2
Technical, Content & Link Diagnostics
Semantic map of the site, JS rendering checked as Googlebot sees it, entity and schema audit, backlink profile filtered for discounted donors. Each finding tied to the pages and queries it affects.
Week 3
Delivery & Handover
The backlog handed over as tickets, the two-page brief to stakeholders, and two weeks of written support open for your engineers.
Audits We Run
One SEO audit method, tuned to how your site sells
Frequently Asked Questions
Before You Ask
Most agency audits are repackaged crawler exports. Send it over. If it already covers link quality, content against the top 10 and rendering, we'll tell you, and you'll save the budget.
You should, and you’ll get the checklist layer — the technical SEO audit guide walks through it, with the reports and filters named. What a crawler can’t see is which links Google has stopped counting, what your page adds beyond the top 10, and whether a category has a hub page at all. That’s the layer we built Algorithm for.
A person. Algorithm runs the scans, but the reading of them is done by Roman and one senior specialist per engagement, and each finding comes with the reasoning written out. Nothing in the report is a tool's verdict passed on unread.
All three. Technical SEO: crawlability, indexing, JS rendering, Core Web Vitals and page speed, schema markup. On-page: content quality, intent match, internal linking, keyword cannibalization and what the page adds beyond the top 10. Off-page: the backlink profile, toxic and discounted links, anchor distribution. What we don't do is hand them over as three separate checklists. The fee is earned where the three overlap, and that's usually where the drop sits. The clearest case is a B2B site whose rankings are fine and pipeline isn’t — how that audit reads the join between Search Console, GA4 and the CRM is written up in why your traffic isn’t reaching the pipeline.
Engagements start at $1,500 and are quoted as a fixed scope once we've seen the site. That sits at the low end of what agencies charge for a manual audit, and the reason is not a smaller scope: Algorithm does the data collection, so the budget goes to the two people reading it.
Then you get a documented confirmation that the architecture is sound, which is what you want in hand before a migration, relaunch or acquisition. On a large site that confirmation runs template by template rather than page by page; the enterprise SEO audit guide shows what it checks.
Read-only Google Search Console and analytics. Server logs help but aren't required. No code access needed for the diagnostic phase.
Then we say so and show the numbers: what patching would cost against what a rebuild would. Design, development and SEO sit in-house, so a rebuild can go from architecture to launch with the same people who ran the audit.
You talk to the person doing the work
Direct Contact
Lead at Neon. Runs its SEO audit services and built Algorithm, our in-house analytics platform. 10+ years in search & digital design.
Roman Makuev
Not sure if your website needs an SEO audit?
Send us your URL. We'll take a first look and tell you whether an audit makes sense for where the site is now. If it doesn't, we'll say that too.