METHODOLOGY · EVIDENCE · LIMITS
How SocialKit produces trustworthy results.
We separate what a tool directly observes from what it estimates, and we avoid presenting unavailable data as fact. This page explains the evidence hierarchy used across SocialKit.
Our evidence hierarchy
Inputs supplied by the user, browser-calculated values, or fetched public-page evidence such as titles, headings, links, schema and robots directives.
Metrics read from an authorized provider such as Search Console or a licensed keyword/backlink/SERP data source when the account is actually connected.
Rule-based prioritization built from observed evidence. These scores are guidance, not Google scores, traffic estimates or market-share facts.
What website-audit tools can and cannot know
| Signal | What we can verify | What we do not claim without a source |
|---|---|---|
| HTML / SEO structure | Fetched title, H1, headings, meta, canonicals, links, schema and selected technical signals. | That every page on the entire site has the same condition unless full crawl coverage is actually verified. |
| Competitor research | Exact-niche verification from fetched evidence, bounded topic-focused samples, confirmed failures and repeated supported phrases. | Private traffic, conversions, all rankings, backlinks or a traffic-ranked “Top 10” without connected/licensed data. |
| Search performance | Real clicks, impressions, CTR and average position when an authorized Search Console property is connected. | Search Console metrics when the property is disconnected. |
| Keyword / backlink / SERP data | Provider-returned values when a licensed source is configured. | Volume, difficulty, referring domains, rankings or AI-answer presence when no provider is connected. |
Bounded crawl disclosure
Interactive website audits use a bounded public-HTML sample to stay fast and respectful of remote servers. A sampled result is evidence from the pages actually fetched; it is not automatically a full-site crawl. GapForge can use 5, 10 or 15-page interactive topic-focused samples, while saved-project monitoring can sample more public pages. JavaScript-rendered sites can require an authorized render bridge.
Strict niche trust in competitor analysis
GapForge does not treat broad similarity as enough. A discovery candidate is only a hypothesis. Fetched evidence must verify the same exact sub-niche and a compatible website model before competitor gaps can unlock. Ambiguous evidence stays unverified; a proven mismatch is blocked. This prevents unrelated “gap” recommendations from being shown as real opportunities.
Privacy and data handling
Browser-only utilities process data locally where practical. Tools that need public website evidence or account/provider data use server-side requests required for that workflow. See the Privacy Policy for current data-handling details. Do not send passwords, card numbers, CVV, UPI PINs or banking credentials through support forms or tool inputs.
No fake proof policy
Who operates SocialKit
socialkit.tools is built and maintained by M/s. Universal Hospitality, Navi Mumbai, Maharashtra, India. Contact routes and billing-support guidance are published on the Contact page. Usage, privacy, disclaimer and refund terms are available in the global footer.
How we update trust content
- Correct stale counts, dates or product descriptions when the catalogue changes.
- Keep methodology language aligned with the real implementation.
- Use page-type/schema markup only when visible content supports it.
- Do not turn estimates into factual claims after a UI redesign.
- When a result is blocked, incomplete or ambiguous, preserve that uncertainty in the UI.
Open the Universal Website Trust-Building Prompt → · About SocialKit → · Contact →