1. Start with the right question
Enter a public website or landing-page URL in Website Growth Audit. Choose organic discovery, leads, sales or useful content coverage. The selected goal changes the measurement plan; it does not change the observed check score.
Use a 5-page sample for a quick check or request up to 10 or 15 pages. The server stops at its time/request budget. Fewer pages can be returned when the site is slow, restricted or has few discoverable links.
2. Read the evidence before the score
The sample check score is the weighted percentage of completed pass-or-fail checks. Unknown and review checks are excluded. Weights are SocialKit product heuristics, not Google ranking weights. A score of 100 does not mean the site is error-free, authoritative or ready to rank.
Every finding includes the exact page, the observed signal, a recommended action and a verification step. “Confirmed issue” means a check failed; an intentional noindex on a private-use public page may be appropriate. “Review needed” means a person must interpret the evidence.
3. Prioritize the actual blockers
Use the priority filters to review P1 checks first: unreadable pages, crawl restrictions or search exclusion directives. Verify that the page should be public and searchable before changing these controls. Then review titles, headings, schema syntax, mobile viewport and page-specific descriptions.
Canonical differences, repeated titles, missing alt attributes and little source text need context. An empty alt attribute can be correct for a decorative image. A short utility page does not need filler paragraphs to satisfy a word-count target.
4. Treat links and competitor clues as candidates
Internal-link suggestions come from shared terms in fetched page headings. Open both pages and confirm that a link would help the reader. The tool does not render JavaScript links or establish complete sitewide click depth; a sampled inbound count of zero does not prove an orphan page.
An optional competitor comparison reads one submitted page. Terms missing from your bounded sample may exist elsewhere on your website. Validate relevance, existing coverage and actual demand before commissioning content.
5. Save a baseline, make a change, rerun
Export the JSON report before making changes. After deployment, rerun the same URL and sample settings, then import the earlier JSON in Compare audits. Previously failed checks are marked resolved only when the same page and check now pass. Missing or unvisited checks remain unverified.
Comparisons happen in the browser. JSON and CSV downloads are local files, not a cloud report history or automatic website backup. Existing SocialKit plan protections apply to results and exports.
6. Measure growth separately
For organic discovery, record Search Console clicks, impressions, queries and landing pages over comparable periods. For leads or sales, verify the actual form or checkout journey and record completed conversions. The audit does not connect to analytics, estimate demand or fetch rankings.
Server fetch time is not LCP or INP. Use the dedicated Core Web Vitals workflow and browser testing for performance. Use PillarMesh for topic architecture and RankCraft for a one-page implementation blueprint.
7. Automated Keyword Research
After an audit, open Keyword research in the report sidebar. The workspace suggests up to five distinct phrases from fetched titles and headings. Review and edit these candidates; they are not Search Console queries. Select country, history period and category, then press Find trending keywords.
The public Google Trending Now RSS feed can identify exact keyword listings and related news candidates with lexical overlap. It has its own publication dates and does not use your history period or category. No feed match means unknown, not “not trending.” Worldwide has no feed in this integration.
When the owner connects SerpApi, authenticated Pro users can retrieve normalized interest history and related queries for each seed. SocialKit compares equal recent windows and labels rising/falling only above a 20% and 5-point change. Low or missing data stays insufficient/unknown. Related Rising and Breakout labels come from the source; Top related is not growth. Nothing here is monthly search volume.
Review relevance and source links before using Add to content plan. Coverage checks sampled titles/headings only. Keep useful evergreen pages even during a falling period. Export keyword CSV, content-plan CSV or the full audit JSON, which includes the research. CSV opens in Excel or can be imported into Google Sheets. There is no automatic publishing, scheduled background run or direct Google Sheets connection.
Source documentation: Google Trending Now · SerpApi Google Trends. Hosting setup is in INSTALL_v222.txt in the update archive.
When no report can be verified
If the site returns an HTTP error, challenge page, robots restriction or incomplete response, investigate access first. Unavailable evidence must not be interpreted as poor content. The tool reports those limits and does not invent a traffic or quality score.