Everything you need to use Rank.

Set up Search Console, ask Rank Advisor what to do next, understand every workspace, publish safely, and learn from the history Rank builds locally. This guide covers Rank 1.5 build 8.

Connect GoogleOAuth, properties and syncing Use Rank AdvisorAsk questions and run reviewable actions Publish to WordPressApplication Password setup Use with ElementorSafe manual implementation

No help sections matched that search. Try a platform name, feature, or error.

Getting started

Rank is a local-first macOS workspace. The normal flow is: connect Google Search Console, select a property, sync, crawl the site, review opportunities, approve changes, then export or publish only the work you choose.

Requirements

macOS 15 or later.
Access to at least one Google Search Console property for live performance data.
Network access for Google APIs, crawling public pages and optional publishing integrations.
macOS 26 and a supported Apple Intelligence Mac only if you want on-device Apple Intelligence drafting. All core SEO features work without it.

First ten minutes

Open Rank and choose Sign in with Google.
Approve read-only Search Console access in your browser and return to Rank.
Choose a site from the property selector. Domain and URL-prefix properties are supported.
Run the first sync. Rank loads query, page, country and device performance into its local cache.
Open Site Audit, confirm the starting URL and page limit, then crawl.
Review Opportunities and Optimization Queue. Approve only the changes you intend to implement.

Demo modeYou can explore Rank without connecting Google. Demo data is useful for learning the interface, but monitoring and site-specific recommendations need live data.

Google Search Console

Connect your account

Click Connect in the toolbar or open Settings and choose Connect Google.
Your default browser opens Google's sign-in and consent screen.
Select the Google account that owns or can access the required properties.
Approve read-only Search Console access. Rank cannot edit your site or Search Console settings through this permission.
Return to Rank. Your sites appear in the site selector.

Select and sync a property

Use the globe control in the toolbar to switch properties. The selected property controls all performance views, opportunities, monitoring, audit impact scoring and exports. Choose the default sync window in Settings and use Sync whenever you want a fresh snapshot.

Rank caches successful snapshots locally. Historical snapshots power comparisons and regression detection, so the app becomes more useful after repeated syncs.

What Rank imports

Clicks, impressions, click-through rate and average position.
Query and page dimensions.
Country and device dimensions.
Available Search Console properties and permission levels.

Search Console delayGoogle Search Console is not real time. Recent dates may be incomplete, and position is an average rather than a live rank check.

Disconnect or clear local data

Open Settings to disconnect Google or clear the local Search Console cache. Disconnecting removes Rank's saved Google credentials. Clearing the cache removes local performance history and reduces the historical context available to Compare and Monitoring.

Workspace guide

Workspace	What it does	Best next action
Overview	Summarizes current performance, movement and high-value opportunities.	Confirm the selected site and data freshness.
Rank Advisor	Answers questions using local Rank data and prepares reviewable action cards.	Ask what changed, what to fix, or what deserves attention next.
Portfolio	Shows comparable metrics across pinned Search Console properties.	Pin the sites you check every week.
Compare	Compares properties and historical periods for clicks, impressions, CTR and position.	Use the same date window across properties.
Pages	Lists page-level Search Console performance and related signals.	Look for high impressions with weak CTR or declining clicks.
Opportunities	Turns performance patterns into prioritized SEO and content ideas.	Open the evidence before committing work.
Optimization Queue	Creates reviewable implementation tasks and suggested values.	Edit, approve or skip each item.
Publishing	Previews approved changes against a connected destination.	Confirm the target and exact operation before publishing.
Proof of Impact	Measures published changes at 7, 14, and 28-day checkpoints.	Mark exported or manually deployed work live after deployment.
SEO Timeline	Combines changes, crawls, Search Console snapshots, experiments, and alerts.	Filter the history by event type when investigating a movement.
Experiments	Compares original and proposed title or meta variants using CTR.	Deploy Variant B, sync later, then measure after enough impressions.
Intelligence	Shows the daily briefing, forecasts, decay calendar, templates, and AI referrals.	Start with Briefing, then open the evidence behind a recommendation.
Client Mode	Stores approvals and creates a read-only presentation or HTML report.	Set the client identity before presenting or exporting.

Rank Advisor

Rank Advisor is a local assistant for the selected site. It reads the data Rank has already loaded on your Mac and turns it into plain-English answers with evidence and reviewable actions.

What Advisor can read

Search Console clicks, impressions, CTR, position, queries and pages.
Opportunities, Optimization Queue items, experiments and Proof of Impact records.
Site Audit summaries, crawl issues, indexation signals, redirects, links and on-page findings.
SEO Timeline events, monitoring alerts, daily briefing, forecasts, content decay and template findings.
AI Visibility snapshots, GEO Audit items and Brand Footprint context.
Publishing status for approved work and connected destinations.

Good questions to ask

What should I work on next for this site?
Why did traffic change and what should I check?
Find my best CTR opportunities.
Which crawl issues matter most?
What can improve AI search visibility?
Write a client-ready weekly summary.
Which title or meta tests should I run?
What changed since the last crawl?
Which approved fixes are ready to publish?
Where are my biggest indexation risks?
What pages are decaying this month?
What should I do for ChatGPT or Perplexity visibility?

Action cards

Some Advisor replies include action cards. These can open the relevant workspace, sync Search Console, run a fresh crawl, approve top queued fixes, start metadata experiments, pin the selected site to Portfolio, or create an AI Visibility prompt set.

Review before publishingAdvisor does not silently publish live changes. Publishing still requires approved Optimization Queue items, a selected destination, and your explicit preview and publish step.

Privacy

Rank Advisor in version 1.5 uses local Rank data and deterministic analysis. It does not add another external AI API, cloud assistant, account, or Google permission. If future versions add optional external AI providers, they should be documented separately and require explicit setup.

Site Audit

The crawler runs locally on your Mac. It discovers pages, reads important HTML signals and builds a repair list without uploading the crawl database to Rank servers.

Run a crawl

Choose Site Audit and check the start URL.
Set the page limit. Start smaller on a new or very large site.
Optionally enable capped external-link checks. These add time and requests to other domains.
Start crawling. The progress indicator reports discovered and completed pages plus elapsed time.
Review the specialist Audit sections rather than treating every issue as equal.

Audit sections

Indexation: sitemap membership, indexable pages, noindex directives, canonicals and robots blocking. Google URL Inspection may provide additional status where available.
Links: broken internal and external links, redirect destinations, multi-hop chains, loops, temporary redirects, HTTP-to-HTTPS hops, orphan and weakly linked pages.
On-Page: titles, descriptions, headings, OpenGraph, image alt coverage, schema, thin content and duplicate or near-duplicate content.
Performance: lightweight page signals and optional PageSpeed data. PageSpeed availability depends on Google's quota.
Reports: export and compare crawl snapshots to identify new, fixed and persistent issues.

Crawl etiquette

Rank respects robots signals, caps work, and is designed for sites you own or are authorized to review. Keep external checks opt-in, avoid aggressive repeated crawling, and use a sensible page limit.

Impact prioritizationWhen Search Console data is available, Rank combines crawl findings with clicks and impressions so work affecting valuable pages rises above low-impact housekeeping.

Opportunities

Rank looks for queries near stronger positions, high-impression low-CTR results, decaying pages, emerging topics, useful clusters, recommendation-page formats and AI visibility gaps. Each opportunity includes evidence, estimated impact and a recommended action.

How to review an opportunity

Check that the query intent matches the current page.
Inspect impressions, clicks, CTR and position rather than relying on a label alone.
Decide whether to improve the existing page or create a distinct page.
Move implementation work into the Optimization Queue or Content Studio.

Do not optimize purely for CTRA more compelling title must still accurately describe the page. Misleading metadata may increase short-term clicks but damage trust and conversions.

Optimization Queue

The queue is the control point between analysis and implementation. Rank may suggest titles, descriptions, OpenGraph fields, schema, internal links, answer blocks, FAQs, entity summaries, trust signals, alt text, content refreshes and llms.txt.

Status and review

Queued: suggested but not authorized for publishing.
Approved: eligible for Publishing and implementation exports.
Skipped: intentionally excluded.

Edit suggested values before approval. Entity-based metadata includes the primary entity, supporting entities, search intent and a CTR rationale. Treat these as a starting point informed by current performance, not an automatic guarantee.

Supported does not mean universal

The Publishing preview recalculates support for the chosen destination. For example, a title may be safe for WordPress core while a meta description belongs to a WordPress SEO plugin and must be implemented manually.

Proof of Impact, GA4 and Tinylytics

Proof of Impact closes the loop between publishing a change and learning whether it helped. Rank stores the exact operation, captures a baseline, and evaluates comparable Search Console, crawl, and optional GA4 or Tinylytics data after 7, 14, and 28 days.

How records are created

Live WordPress updates, Shopify product SEO updates, and published Webflow changes start measurement immediately.
WordPress drafts, Webflow metadata-only saves, Framer handoffs, and local HTML changes are recorded as awaiting live.
After deploying an awaiting-live change, open Proof of Impact and choose Mark Live. Rank captures a fresh baseline and schedules its checkpoints.

What Rank captures

Exact before-and-after title, metadata, or implementation values from the publishing preview.
Search Console clicks, impressions, CTR, and average position for the affected page.
The number of current crawl issues associated with that page.
When GA4 is connected: organic landing-page sessions, key events, and reported revenue.
When Tinylytics is connected: page views, unique views, Kudos, and recognized AI referral sources.

How outcomes are classified

A checkpoint becomes a win when clicks, key events, revenue, or crawl health improve meaningfully; a loss when they regress meaningfully; otherwise it remains inconclusive. Low-volume results include a caution because small movements can be noise.

Not causal proofRank compares like-for-like reporting windows, but seasonality, campaigns, algorithm updates, site changes, and competitors can also influence results. “Approximately X additional clicks” is an estimate, not a controlled experiment.

Automatic rechecks

While Rank is open, it watches the selected site's next due checkpoint. At 7, 14, and 28 days it refreshes Search Console plus connected GA4 or Tinylytics data, then records the result. If the Mac or Rank is closed at the due time, the next sync or crawl captures overdue checkpoints.

Connect Google Analytics 4

Open Proof of Impact or Settings and choose Connect GA4.
Sign in with a Google account that has access to the required Analytics property.
Approve the separate read-only Google Analytics permission.
Select the GA4 property that corresponds to the active Search Console site.
Choose Sync GA4. Rank loads organic-search landing pages for the same reporting window used by Search Console.

GA4 data also affects Optimization Queue priority: pages with existing organic key events or revenue receive a measured business-impact boost.

Developer setup required for GA4

The Rank Google Cloud project must have both Google Analytics Data API and Google Analytics Admin API enabled. Add https://www.googleapis.com/auth/analytics.readonly to OAuth Data Access. Google categorizes scopes in Cloud Console; if it is shown as sensitive, provide the required justification and submit the changed app for OAuth verification before public release.

Separate authorizationGA4 uses a separate Keychain token from Search Console. Users who do not connect GA4 keep the existing Search Console-only workflow.

Connect Tinylytics

Open Tinylytics and create or copy a read-only API key from Account Settings → API Access.
Open Rank → Settings → Tinylytics.
Paste the read-only key and choose Connect Tinylytics.
Select the Tinylytics site that matches the active Search Console property.
Choose Sync Tinylytics. Rank loads page views, unique views, Kudos, and grouped source/referrer signals for the current reporting window.

Tinylytics is optional and uses a separate API key stored in Keychain. It is useful when you want lightweight first-party analytics in Rank without enabling another Google API. Tinylytics views and Kudos can influence Optimization Queue priority; GA4 is still the source for reported revenue.

Rollback and restoration

Every record keeps previous values and destination-specific rollback guidance. Local HTML changes include a real Restore Original HTML action. Rank preserves the currently deployed file as a new safety copy before restoring the captured original.

Timeline, Experiments and Rank Intelligence

Rank's intelligence layer turns the local history created by syncs, crawls, publishing and monitoring into a practical decision system. These features do not require a new account, backend, OAuth provider or Google scope.

SEO Change Timeline

The Timeline is the investigation log for the selected Search Console property. It merges published optimization records, completed crawls, stored Search Console snapshots, experiments and monitoring alerts in reverse chronological order.

Select the property you want to investigate.
Open SEO Timeline.
Use the event-type menu to isolate Changes, Crawls, Search, Alerts or Experiments.
Compare the dates around a ranking or traffic movement. Open the relevant source workspace when you need its full evidence.

History starts locallyRank cannot reconstruct events from before they were recorded on this Mac. Repeated syncs, crawls and publishing actions make the timeline more useful over time.

Run a title or meta experiment

Experiments are available for title and meta-description recommendations already generated in Optimization Queue. Rank records the original as Variant A, the proposed value as Variant B, and the current page-level Search Console metrics as the baseline.

Sync Search Console with the reporting window you intend to keep using.
Open Experiments and choose a title or meta-description candidate.
Choose Start Experiment to capture Variant A, Variant B and the baseline.
Publish Variant B through Rank or your normal CMS workflow. Rank deliberately does not switch live metadata merely because an experiment was created.
Wait for Google to recrawl the page and for a comparable Search Console window to accumulate.
Sync again, return to Experiments and choose Measure Now.

Rank reports Variant B won, Original won, or Inconclusive. Fewer than 100 impressions is treated as insufficient volume. At higher volume, Rank looks for movement beyond a small materiality threshold rather than calling every decimal change a winner.

Not a split testSearch engines do not provide simultaneous A/B delivery here. Seasonality, ranking changes, snippets rewritten by Google and other site work can affect CTR. Keep the reporting window consistent and treat the result as strong directional evidence.

Daily Native Briefing

Open Intelligence → Briefing for the three highest-priority locally recorded events from the previous 24 hours. High-severity alerts appear before routine syncs and changes. When nothing needs attention, Rank shows a quiet caught-up state instead of inventing work.

Opportunity Forecasting

Forecasts estimate additional clicks for queued work using current impressions and Rank's impact/confidence score. When optional analytics are connected, Rank can use GA4 conversions/revenue or Tinylytics page activity and Kudos to better rank work by likely outcome.

Without GA4, click forecasts still work; conversion and revenue estimates remain zero.
Without Tinylytics, lightweight page-view and Kudos signals remain unavailable.
Forecasts rank the queue before implementation. Proof of Impact measures what happened after publishing.
Forecasts are not financial promises and should be reviewed alongside confidence, evidence and effort.

Content Decay Calendar

The calendar converts pages currently classified as declining or needing work into an ordered refresh list. Each row includes a suggested review date, risk score, approximate clicks at risk and the reason Rank scheduled it. Refresh dates are decision aids rather than crawling or publishing schedules.

Template Intelligence

After a crawl, Rank groups repeated URL shapes such as product or article families. A finding appears when a pattern has at least three crawled pages and the same class of issue affects multiple members. Use the common issue counts and sample URLs to decide whether one WordPress template, Shopify theme template or shared component is the true repair point.

Detection, not automatic theme editingRank infers patterns from URLs and crawl evidence. It does not claim to know the internal CMS template and does not modify theme or template code automatically.

AI Referral Analytics

When GA4 or Tinylytics is connected, Intelligence → AI Referrals classifies recognized sources for ChatGPT, Perplexity, Gemini, Copilot and Claude. GA4 can display sessions, key events and reported revenue; Tinylytics can display visits and Kudos from recognized grouped sources/referrers.

Referral attribution is not complete: browsers, redirects, privacy controls and AI products can omit or alter referrer data. An empty result means the connected analytics source returned no recognized source for the selected window, not necessarily that no visitor used an AI assistant.

Client Mode and approvals

Open Client Mode.
Enter the client name, report title, prepared-by line and optional accent value.
Choose an approval state for recorded changes: Pending, Approved or Changes requested.
Choose Present to open a separate read-only progress window for a call or screen share.
Choose Export HTML for a portable branded report containing headline metrics and recent progress.

Approval states and branding settings stay in Rank's local cache. HTML reports contain the report content you chose to export; review the file before sharing it externally.

Content, AI Visibility and GEO

Content Studio

Turn opportunities into ideas, briefs, outlines, drafts, titles, descriptions and FAQs. On macOS 26 with supported hardware, Rank can use Apple Intelligence for enhanced on-device drafting. On macOS 15, Rank clearly falls back to deterministic local drafts. Always review factual claims, citations and brand voice.

LLM Search and AI Visibility

Record prompts and responses from AI answer engines, then track brand mentions, citations, linked sources, competitors, sentiment and share of voice over time. AI answers vary, so compare repeated snapshots rather than treating one answer as a stable ranking.

GEO Audit

Review crawlability for AI crawlers, entity clarity, direct-answer structure, source and trust signals, schema, freshness, extractability and llms.txt. Rank can generate recommendations and implementation snippets; it cannot guarantee inclusion in an AI answer.

Brand Footprint

Identify listicles, review sources, comparison pages and competitor mentions that may influence both buyers and AI systems. Use outreach targets responsibly. Rank is not an automated posting or spam tool.

Competitors and sitemap analysis

Crawl your own site first for the strongest comparison.
Open Competitors and add one authorized public homepage, category, article, product or sitemap URL per line.
Run Gap Analysis. Rank uses capped samples and public sitemap patterns.
Review recurring URL structures, content types, recommendation pages and missing topic ideas.
Use the result as research. Do not copy competitor content or proprietary material.

Recommendation Page Finder can suggest useful “best”, “alternatives”, “versus” and audience-specific pages based on real gaps. Publish these only when you can provide honest criteria, evidence and a genuinely useful comparison.

Monitoring and native alerts

Rank compares each new local sync or crawl with a previous snapshot. It can generate in-app alerts and native macOS notifications for ranking, traffic, crawl, indexation and AI visibility changes.

Open Monitoring and choose Enable All, or enable individual channels.
Allow notifications when macOS asks. If denied, use System Settings → Notifications → Rank.
Add important queries and pages to the watchlist.
Choose Sensitive, Balanced or Calm presets to set practical regression thresholds.
Sync and crawl on a recurring cadence. Rank needs a previous snapshot before it can detect change.

Rank is not a background cloud monitorThe app is local-first. Comparisons happen when Rank receives a new sync, crawl or saved AI snapshot; they are not guaranteed while the Mac is off or the app cannot run.

Reports, exports and Share Cards

Audit Reports can export crawl findings and compare historical snapshots. Implementation exports include metadata CSV, internal-link snippets, JSON-LD, notes and content briefs. Open exported files before sending them to a client and confirm that the selected property and date range are correct.

Share Cards create branded PNG summaries for traffic wins, crawl progress, CTR opportunities and AI visibility. Cards use the Rank icon plus Rank and yuzool.com branding while keeping the selected site's progress central.

Publishing overview

Publishing never starts from raw recommendations. Approve changes in Optimization Queue, choose the page URL and destination, connect the destination, inspect the preview, then publish or export.

Approve one or more Optimization Queue items.
Open Publishing and select the relevant page URL.
Choose WordPress, Webflow, Shopify, Framer or HTML.
Connect or select the target. Rank attempts to match the approved URL.
Read every preview row. Green rows are supported; orange rows explain why Rank will not apply that operation automatically.
Publish only after confirming the target, before/after value and destination mode.

Keep your own rollback pathRank creates automatic backups for local HTML. For CMS platforms, also use the platform's revisions, backups, staging or draft workflow before material changes.

Destination	Automatic support in 1.5	Deliberately excluded
WordPress	Core title; selected snippets appended to non-Elementor content; draft-copy option.	SEO-plugin fields, full content rewrites, automatic Elementor layout edits, media matching.
Webflow	Page SEO title, description and OpenGraph metadata; optional page publish.	Designer layouts and CMS collection content.
Shopify	Product SEO title and description.	Themes, templates, product copy, pages and checkout.
Framer	Credential-free JSON and Markdown handoff export.	Direct live publishing while Framer's Server API requires its beta Node runtime.
Local HTML	Title, description, OpenGraph, schema and selected snippets with backup.	Ambiguous image-alt changes, broad content rewrites and automatic `llms.txt` placement.

WordPress setup and Application Passwords

Rank uses WordPress's built-in REST API and a dedicated Application Password. Do not enter your normal WordPress account password.

Create the Application Password

Sign in to WordPress Admin over HTTPS.
Open Users → Profile, or Users → All Users and edit the account Rank should use.
Scroll to Application Passwords.
Enter a label such as Rank for Mac and choose Add New Application Password.
Copy the generated password immediately. WordPress only displays it once.
In Rank → Publishing → WordPress, enter the public site URL, the WordPress username and the generated Application Password.
Choose Connect. Rank validates the account and loads editable posts and pages.

How Rank stores itThe Application Password is stored in this Mac's Keychain using device-only accessibility. Revoke it any time from the WordPress user profile.

Recommended permissions and account

Use a dedicated WordPress user with only the role needed to edit the intended posts or pages. Do not use an Administrator account when an Editor account is sufficient. Use HTTPS and keep WordPress updated.

Update existing or create a draft

Update existing: applies supported changes to the selected resource.
Create draft copy: creates a separate unpublished copy titled with “Rank draft” for review. Prefer this for the first use on a production site.

WordPress SEO plugins

Yoast, Rank Math and other plugins store SEO metadata outside WordPress core. Rank does not write those private plugin fields. Copy the approved title or description into the plugin's editor manually when Publishing marks it unsupported.

If Application Passwords are missing

Confirm the site runs WordPress 5.6 or later and uses HTTPS.
Check whether a security plugin, host policy or custom code disabled Application Passwords or REST API authentication.
Confirm you are editing your own user profile and have permission to create credentials.
Ask the host or site administrator before changing security controls.

Using Rank with Elementor

Rank detects likely Elementor-managed content and protects it from automatic body changes. Elementor stores layout data separately, so appending HTML through WordPress core can damage or bypass the visual layout.

What Rank can safely do

Analyze the live Elementor page through Search Console and the local crawler.
Generate titles, descriptions, structured-data suggestions, answer blocks, FAQs, internal-link ideas and content briefs.
Update the WordPress page title when approved.
Show implementation snippets while marking layout-sensitive operations as unsupported.

Implement an approved Elementor change

In Rank, approve the recommendation and open Publishing → WordPress.
Select the matching Elementor page. Orange preview rows identify protected layout work.
Copy the suggested value or implementation snippet from Optimization Queue.
In WordPress, choose Edit with Elementor for that page.
Use a Heading, Text Editor, HTML or relevant widget depending on the recommendation. Add internal links through Elementor's link control rather than pasting arbitrary markup into the layout.
For schema, use the SEO plugin's schema controls or a carefully placed HTML widget if your setup supports it. Avoid duplicating schema already generated by the theme or plugin.
Use Elementor's responsive preview, then Preview Changes in WordPress.
Update the page, clear relevant caches and run a focused Rank crawl to verify the result.

Metadata on Elementor sites

Elementor usually does not own the canonical SEO title and meta description. Edit those fields in the active SEO plugin, such as Yoast or Rank Math, using Rank's approved values. Check the rendered page source after publishing to ensure only one title, description and canonical are present.

Safer rollout

Use staging or a recent site backup for material layout work.
Apply one group of changes at a time.
Check desktop, tablet and mobile breakpoints.
Verify links, forms and structured data after cache/CDN invalidation.

Webflow

In Webflow, create a scoped Site API token for the intended site.
Grant the Pages read/write and Sites read/write access required by Rank.
Paste the token into Rank → Publishing → Webflow and connect.
Select the Webflow site and matched page.
Preview SEO title, description and OpenGraph operations.
Choose whether Rank should publish the matched page after updating metadata.

The token stays in Keychain. Rank does not edit Webflow Designer layouts or CMS collection content in this version. Revoke the token from Webflow if the Mac or credential is no longer trusted.

Shopify

Create a token

In Shopify Admin, create or select a custom app for the store.
Configure Admin API scopes with read_products and write_products.
Install the app on the store and reveal the Admin API access token.
Copy the store's permanent your-store.myshopify.com domain and the token into Rank.
Connect, choose the matched product and preview the exact SEO operation.

Rank updates the product SEO title and description through Shopify's Admin GraphQL API. It does not modify themes, templates, product body copy, regular pages, checkout or inventory. The token is stored in Keychain.

Treat Shopify tokens as sensitiveUse the narrowest scopes, never paste a token into a report or Framer export, and revoke/reissue it if exposed.

Framer

Framer's official Server API currently uses a beta, stateful Node.js 22 client. Rank therefore uses a portable handoff rather than embedding that runtime or pretending to provide direct live publishing.

Approve the required changes in Optimization Queue.
Open Publishing → Framer and enter the Framer project URL for reference.
Choose an export folder.
Review the changes and choose Export for Framer.
Open the generated timestamped folder. It contains rank-framer-changes.json for structured tooling and changes.md for human implementation.
Apply and preview the values in Framer before publishing the project.

No Framer API key is written into the export. When Framer's production API becomes suitable for a sandboxed native app, direct publishing can be added without changing the approved-change workflow.

Local HTML sites

Keep a version-control or deployment backup of the site.
Choose the root folder containing the site's .html or .htm files.
Rank matches the selected page URL to common paths such as about.html or about/index.html. It can also use the canonical URL inside a file.
Review the matched filename and every proposed operation.
Choose Update HTML File. Rank first copies the original into .rank-backups, then writes atomically.
Review the local diff and deploy through your normal hosting workflow.

Supported operations include titles, descriptions, OpenGraph, schema and selected body snippets. Rank refuses ambiguous image-alt changes, broad content rewrites and automatic llms.txt placement.

Privacy and security

For the dedicated Google OAuth and Search Console data disclosure, see the Rank Privacy Policy.

Search Console and GA4 OAuth tokens, Tinylytics API keys, plus WordPress, Webflow and Shopify credentials, are stored in macOS Keychain.
Search Console snapshots, GA4 and Tinylytics metrics, crawl results, Advisor messages, change history, experiments, client approvals, report settings, monitoring state and AI visibility records are stored locally.
Rank Advisor, Timeline, briefing, forecasting, decay, template and Client Mode features add no cloud account, backend or API permission.
Rank does not require a Yuzool account for normal use and does not use developer advertising tracking.
Google access is read-only. Tinylytics should use a read-only key. Publishing credentials are destination-specific and should use minimum permissions.
Framer exports intentionally contain no API credential.
Local HTML writes create timestamped backups, but CMS publishing still relies on the destination's revisions and backups.

Only connect accounts and crawl sites you are authorized to use. Revoke destination credentials when they are no longer needed, before transferring a Mac, or after suspected exposure.

Troubleshooting

Google sign-in completes but Rank does not connect

Return to Rank after the browser confirms authorization. If it still fails, disconnect in Settings, quit and reopen Rank, then sign in again. Confirm the Google account has Search Console access and that local network callbacks are not blocked by security software.

A Search Console property is missing

Verify the exact Google account and property permission in Search Console. Domain and URL-prefix properties are separate. Reconnect or sync after access changes.

WordPress returns 401 or 403

Confirm the WordPress username, regenerate the Application Password, and check that the site uses HTTPS. A host, firewall or security plugin may block REST API authentication. Do not weaken security controls without the site owner's approval.

WordPress connects but an SEO field is unsupported

Meta descriptions and OpenGraph fields commonly belong to Yoast, Rank Math or another plugin rather than WordPress core. Copy the approved value into that plugin manually.

Rank will not edit Elementor content

This is intentional protection. Copy the snippet or value from Optimization Queue and apply it through the appropriate Elementor widget, then recrawl the page.

Shopify does not load products

Use the permanent .myshopify.com domain, not a storefront vanity domain. Confirm the custom app is installed and the token has read_products and write_products.

GA4 connection returns HTTP 403

Confirm that the Google Analytics Data API and Admin API are enabled in Rank's Google Cloud project, the OAuth consent configuration includes analytics.readonly, and the signed-in Google account can access at least one GA4 property.

A GA4 property or landing page is missing

Reconnect with the correct Google account, choose the matching property, and sync again. Rank requests organic-search landing pages for the active reporting window; pages with no qualifying sessions may not be returned.

Tinylytics connection returns 401 or 403

Confirm the API key is copied correctly, starts with a Tinylytics read-only prefix, and still has access to the Tinylytics sites you expect. Reconnect in Rank Settings after rotating the key.

AI Referrals is empty

First confirm that GA4 or Tinylytics is connected, the correct property/site is selected and a sync has completed. Rank only displays sources it can classify as ChatGPT, Perplexity, Gemini, Copilot or Claude. Privacy controls and redirects can remove referral information, so an empty view is not proof that AI assistants sent no visitors.

An experiment is inconclusive

Keep the Search Console reporting window consistent and allow more impressions to accumulate. Rank treats fewer than 100 impressions as insufficient and avoids declaring a winner for small CTR movement. Google may also rewrite a search snippet instead of showing the supplied metadata.

Template Intelligence shows no findings

Run a crawl that includes enough pages from the repeated URL family. Rank requires at least three pages in a pattern and issues on multiple members before suggesting a shared-template cause.

The Daily Briefing says I am caught up

The briefing only considers events Rank recorded during the previous 24 hours. Sync, crawl or keep monitoring active to add new evidence. Rank intentionally does not invent recommendations when no new local event deserves attention.

Rank Advisor gives a short or generic answer

Advisor can only use data already loaded in Rank. Sync Search Console, run a crawl, review the Optimization Queue, and refresh AI Visibility or GEO Audit to give it more local evidence.

No local HTML file matches

Confirm the selected folder is the site's root and the page follows a common filename pattern. Add a correct canonical URL to the HTML or select the matching page URL in Rank.

PageSpeed reports quota exhausted

Google's PageSpeed quota can be unavailable or rate limited. Rank's local crawl and lightweight performance checks still work; retry PageSpeed later.

Notifications do not appear

Open System Settings → Notifications → Rank and allow notifications. Then enable the desired channels in Monitoring. A previous snapshot and a later sync/crawl are required before a regression exists.

Apple Intelligence is unavailable

Enhanced drafting requires macOS 26 and supported Apple Intelligence hardware/configuration. Rank displays a note and uses its local deterministic drafting workflow on macOS 15.

Updates and support

Use the Polar customer portal to access purchases, receipts, invoices and available Rank downloads. Rank → Help → Rank Help opens this guide from the app.

When reporting a problem, include the Rank version/build, macOS version, affected workspace, destination platform, exact error text and the steps immediately before it occurred. Never send passwords, OAuth tokens or API access tokens.

Current documented release: Rank 1.5 build 8 for macOS 15 or later.