Backlinks
Conduct a backlink audit using AI to analyse link profiles, identify toxic links, and get actionable recommendations to strengthen SEO.
Quick start
1. Enter the target domain and click Submit
2. Review the spam score and health metrics at the top
3. Click Analyse with AI for strategic recommendations
4. Add priority actions to Kanban or export a disavow file
Understanding the display

The health header shows key metrics at a glance:
| Element | Meaning |
| Spam Score | Overall link profile risk (0-100). Green is healthy, amber needs attention, red requires action. |
| Total Backlinks | Number of external links pointing to the domain. |
| Referring Domains | Unique websites linking to the domain. Higher diversity is better. |
| Referring IPs | Unique IP addresses. Low diversity may indicate link networks. |
| Broken | Links pointing to pages that no longer exist. Recoverable link equity. |
Spam score benchmarks
| Score | Meaning |
| 0-30 | Healthy profile. Maintain current link building practices. |
| 31-60 | Moderate risk. Review flagged domains and consider selective disavows. |
| 61-100 | High risk. Immediate attention required. Run AI Analysis for priorities. |
What should I do next?
The spam score determines where to focus effort:
| Spam Score | Action |
| 0-30 | Focus on growth. Build more authority links from quality sources. |
| 31-60 | Balance growth with cleanup. Run AI Analysis to identify quick wins. |
| 61-100 | Prioritise cleanup. Export disavow file and address toxic domains first. |
| Rising trend | Investigate recent link sources. Check for negative SEO or poor outreach. |
Exploring the data
Four tabs provide different views of the backlink profile:
Trends shows backlink growth over time. The dual-axis chart displays total backlinks against referring domains. The second chart tracks new versus lost links monthly.
Geography reveals where links originate. A world map highlights countries, while the TLD chart shows domain extensions. Geographic concentration above 70% in one country may limit international reach.
Link Types breaks down link characteristics: text versus image links, dofollow versus nofollow attributes, and semantic locations (article, footer, sidebar). Editorial links in article content carry more weight.
Sources categorises referring sites by platform type: blogs, news sites, forums, social media. A healthy profile has diverse platform representation.
Using the filters
Filters help isolate specific link types for review:
| Filter | Shows |
| Spam ≥ 80 | High-risk links from potentially harmful domains. |
| IP domain | Links from IP addresses instead of domain names. Often suspicious. |
| Broken | Links to pages returning 4xx or 5xx errors. Recovery opportunities. |
| HTTP source | Links from non-secure pages. Minor trust signal. |
| Indirect | Links through redirects. May dilute link equity. |
| Nofollow | Links that do not pass ranking value directly. |
| Rank-gap ≥ 15 | Links where page rank is much lower than domain rank. Quality concern. |
Tick multiple filters to build a focused list. The table updates automatically. Export filtered results as a disavow file using the export button.
The backlinks table

Below the charts sits the detailed table listing every backlink found. Each row represents a single link and includes:
| Column | What it tells you |
| Source URL | The page linking to your site. |
| Anchor | The clickable text used for the link. |
| Target URL | The specific page on your site being linked to. |
| Spam | Risk score from 0–100. Colour-coded: green is safe, red is toxic. |
| Attributes | Tags like Dofollow, Nofollow, Broken, Indirect, Noopener, or Noreferrer. |
| Domain Rank | Authority of the linking website (0–100). Higher is better. |
| Page Rank | Authority of the specific linking page. Often different from the domain score. |
Expanding a row for more detail
Click the > chevron to the left of any row. An expanded panel opens with additional context:
• Country – geographic origin of the linking domain, shown with a flag.
• Link Value – a combined score factoring in domain rank, placement, and relevance.
• Quick facts – platform type, link type, semantic location, and page language.
• Page Title – the title of the page linking to you.
• Link Context – the surrounding text where the link appears. The anchor text is highlighted so you can see exactly how the link sits within the content.
• External/Internal link counts – how many other links that page has, which affects how much authority flows to yours.
• Security warnings – flags if the linking page uses HTTP instead of HTTPS.
|
Why this matters A link from a high-authority page with few outbound links passes more value than one from a low-authority page with hundreds of links. The expanded view helps you spot which links genuinely move the needle. |
AI Analysis
AI Analysis examines the complete backlink profile and delivers specific, evidence-based
recommendations.

Running an analysis
1. Fetch backlink data first using the Submit button
2. Click Analyse with AI
3. Wait for processing (typically 30-60 seconds)
4. Review recommendations organised by action type
What the analysis delivers
Recommendations fall into five categories:
| Category | Purpose |
| Link Removals | Domains to add to a Google disavow file. Includes specific domains and evidence. |
| Anchor Diversification | Adjustments to anchor text distribution. Targets natural patterns. |
| Authority Acquisition | Opportunities to build links from high-quality sources. |
| Broken Link Recovery | High-authority broken links with suggested redirect targets. |
| Profile Monitoring | Domains and patterns to watch for ongoing risk management. |
Priority levels
| Priority | Meaning | Timeframe |
| Immediate | Large gaps on high-volume keywords or high-risk domains | 1-2 weeks |
| Near-term | Moderate gaps or unstable patterns | 1-2 months |
| Can-wait | Minor improvements for optimisation | When time permits |
Adding recommendations to Kanban
5. Find the recommendation to action
6. Click + Add to Kanban
7. Select the project board
8. Task creates with all details pre-filled
Exporting data
CSV Export downloads the complete backlink table with all attributes. Use for detailed analysis in spreadsheet software.
Disavow Export generates a text file formatted for Google Search Console. Tick filters first to select which links to include. The file contains domain-level entries for IP addresses and URL-level entries for standard domains.
Troubleshooting
No data found
The domain may have no tracked backlinks, or the format is incorrect.
• Enter the domain without https:// or www.
• Try the root domain instead of a subdomain
• New domains may not have backlink data yet
AI Analysis button disabled
Backlink data must be fetched before running AI Analysis.
• Click Submit first to load the backlink data
• Wait for the table to populate
• Then click Analyse with AI
AI Analysis taking too long
Large backlink profiles require more processing time.
• Wait up to 2 minutes for large domains
• If processing fails, try again
• Contact support if the problem persists
Chart not displaying
Charts require data to render.
• Switch to a different tab and back
• Refresh the page and resubmit
• Some domains may lack data for specific chart types
FAQ
How often does backlink data update?
Backlink data updates daily. Historical trends reflect data from the past 12 months.
What counts as a toxic link?
Links with spam scores above 70, unusual anchor patterns, broken status, or from IP-based domains. The AI Analysis provides specific domain-level evidence.
Should all nofollow links be removed?
No. Nofollow links are a natural part of any backlink profile. Only consider removal if combined with other risk factors like high spam scores.
How does the disavow file work?
Upload the exported file to Google Search Console. Google will ignore the listed links when assessing the site. This does not remove the links; it tells Google to discount them.
Why do different tools show different backlink counts?
Each tool uses its own crawling infrastructure and index. Differences of 10-20% between tools are normal. Focus on trends rather than absolute numbers.
Getting help
Contact support for persistent data issues