Scaup Product Context

Copy this document and paste it into ChatGPT, Claude, or any AI tool. Then ask it anything about Scaup - pricing objections, competitor comparisons, customer-specific questions.

How to use this

1. Click "Copy to clipboard" below

2. Open ChatGPT, Claude, or any AI assistant

3. Paste the context and then ask your question, e.g.:

"A prospect uses Elementor and wants to know if Scaup works for them. What should I say?"

"Compare Scaup to hiring a freelance SEO person for a 5-person startup"

"A developer is asking how we edit their codebase safely. Give me the technical explanation."

scaup-product-context.md
# Scaup - Complete Product Context for Sales Use this document as context when answering questions about Scaup. It contains everything a salesperson needs to know to handle any customer question, objection, or scenario. --- ## What Scaup Is Scaup is an automated SEO and AI visibility service. You connect your website once, and Scaup improves your search visibility every day - automatically. No dashboards to check, no jargon to learn, no manual work. The customer gets a recap email - either weekly ("Here's what we improved this week") or daily, depending on their preference. That's the entire experience for most users. Behind the scenes, Scaup makes one improvement per day. The customer picks how often they want to hear about it. They never need to log into a dashboard, configure settings, or learn any SEO terminology. Scaup is NOT an SEO tool (like Semrush or Ahrefs). Those tools give you reports and expect you to do the work. Scaup IS the work. It's the difference between a fitness tracker and a personal trainer. --- ## The Market Opportunity AI tools (Bolt, Lovable, v0, Cursor, Wix AI, etc.) have made it trivially easy to build a professional website. Millions of small businesses, indie hackers, and freelancers now have a site. But having a website and being found on Google are two different things. Most of these sites have zero organic traffic because they were never optimized for search. Scaup is the natural next step: you built your website with AI, now promote it with AI. This is also why "SEO is dead" is wrong. Google still drives 90%+ of search traffic. AI search (ChatGPT, Perplexity, Google AI Overviews) is growing but it's additive - and AI tools cite the same well-structured content that ranks on Google. Optimizing for one optimizes for both. --- ## Who It's For **Primary audiences:** - Indie hackers and solo founders who built a site but have zero Google traffic - Small startups (2-10 people) without a dedicated marketing person - Developers with side projects who'd rather code than do SEO - Freelancers and small agencies managing client websites - Anyone who built a site with AI tools and needs it to actually be found - Local businesses with a website that doesn't show up when people search for what they do - People who tried SEO tools like Semrush, saw 200 "issues," felt overwhelmed, and never came back **What these people have in common:** They know their site needs to be found on Google. They don't have time to learn SEO. They don't have budget for an SEO agency. They want someone (or something) to just handle it. **Who it's NOT for:** - SEO professionals who want granular manual control over every keyword and meta tag - Large agencies needing white-label dashboards, bulk operations, or client portals - Enterprise companies requiring SOC2 compliance, SLAs, or dedicated account managers - Anyone who expects guaranteed #1 rankings in a week (that's a scam, not a service) - Sites with zero content at all (we optimize what exists and create new content, but there needs to be a starting point) --- ## How It Works - Detailed ### Setup (under 5 minutes) 1. **Sign up with Google** - one-click OAuth, same Google account they already use 2. **Connect their site** - pick Lovable or GitHub and select a repository, enter WordPress site URL + credentials, connect via OAuth on Webflow, or install the Scaup app on Wix 3. **Connect Google Search Console** - one click (same Google account, permissions carry over automatically) That's the entire setup. No configuration screens, no settings to tweak, no onboarding wizard. ### What Happens Immediately After Setup - Scaup crawls up to 100 pages on the site (using the sitemap, Google Search Console data, and internal links), recording the structure and content of each one - Scaup identifies the site's top competitors based on overlapping keywords and niche - A competitor gap analysis runs: what do competitors rank for that this site doesn't? - A growth plan is generated automatically with 16+ actionable items, prioritized by impact and cost - For pages with multiple sections, Scaup creates separate improvement items per section so each edit is small and targeted - The plan generates ~50 items for the month, executed one per day so improvements happen gradually ### The Daily Execution Loop (Every day, 6am UTC) 1. **AI picks 1 item**: From the prioritized growth plan backlog, Scaup selects the single highest-impact task for today. This could be a title rewrite, content improvement, technical fix, or new page. 2. **AI execution**: Rewrites the title, improves the content, creates the new page, or updates the technical file. 3. **Validation**: Every change goes through 3 safety layers before applying (details in Safety section). 4. **Apply changes**: Commits to GitHub or pushes via WordPress REST API. 5. **Repeat tomorrow**: The next day, Scaup picks the next highest-priority item and does it again. ### Why One Item Per Day (Not Everything at Once) - **Google needs time to crawl**: Publishing 5 pages at once means they compete for Google's attention. One at a time lets each page get properly indexed. - **Not everything is a new page**: Most days, the highest-impact move is a title rewrite, a content improvement, or a technical fix. Scaup picks whatever moves the needle most, not just new content. - **New content is spaced out**: New pages have a 2-day minimum gap between them. This prevents the site from looking like it's mass-producing content (which Google watches for). - **Monthly quotas scale with site size**: Small sites (under 30 pages) get up to 4 new pages/month. Medium sites (30-100 pages) get up to 8. Large sites (100+ pages) get up to 12. Growth looks organic, not artificial. - **Variety matters**: A site that only publishes new blog posts every day looks one-dimensional. A site that gets title rewrites on Monday, a content improvement on Tuesday, a technical fix on Wednesday, and a new page on Thursday looks like it's being actively maintained by a real team. ### The Recap Email (Daily or Weekly - Customer's Choice) Each site gets a summary email at the cadence the customer picks: daily (a quick "here's what we did today") or weekly (a fuller "here's everything we did this week" on a designated day). Either way, the email covers what Scaup did - pages improved, new content created, keyword movements. Plain English, no jargon, takes 30 seconds to read. Recap emails are automatically translated to the site's primary language. If the site is in Spanish, Portuguese, French, Hebrew, etc., the email arrives in that language - no setting to configure. ### Monthly Competitor Refresh (1st of each month) - Re-runs the competitor analysis to find new keyword gaps - Adds fresh items to the growth plan - Skips items that are no longer relevant (e.g., the site now ranks for that keyword) - Keeps the plan evergreen and responsive to market changes ### The Growth Plan The growth plan is the central concept. It's NOT an audit or a list of issues. It's a prioritized backlog of specific improvements: **Types of plan items:** - **Title/description rewrites** - rewrite a page's title tag and meta description to target specific keywords and improve click-through rate - **Content improvements** - rewrite or enhance sections of an existing page to better match search intent - **New page creation** - create an entirely new page targeting a keyword gap identified from competitor analysis - **Technical file updates** - create or update robots.txt, sitemap.xml, llms.txt, structured data - **Minor content edits** - small targeted changes like adding alt text, improving a heading, tweaking a paragraph Each item has an estimated cost, target keyword(s), and expected impact. The daily execution picks from this backlog based on priority, one item at a time. --- ## What Scaup Does Automatically - Detailed ### Page Titles and Meta Descriptions - The title tag is the single most impactful on-page ranking factor. It's what appears as the blue link in Google results. - The meta description appears below the title in results. It doesn't directly affect rankings but heavily affects click-through rate (whether someone clicks your result vs a competitor's). - Scaup rewrites both for every page, targeting specific keywords identified from GSC data and competitor analysis. - Example: A page titled "About Us" becomes "Austin Yoga Studio | Classes for All Levels | [Studio Name]" - because people search for "Austin yoga studio," not "about us." ### On-Page Content Optimization - Rewrites headings (H1-H6) to include target keywords naturally - Improves paragraphs and sections to better match what people are actually searching for (search intent) - Adds relevant information that searchers expect to find on the page - Removes fluff and filler that dilutes relevance - Preserves the site's voice and factual information - never invents or fabricates - No keyword stuffing - content reads naturally and is genuinely helpful ### New Content Creation - The competitor gap analysis identifies searches where competitors rank but the customer has no page at all - Scaup creates new pages to fill those gaps - New pages are built with a clean, consistent design that reads well on any WordPress setup (custom themes, Gutenberg, Elementor, WooCommerce). The design inherits the site's fonts so the page blends in, and needs no page builder or template - WordPress site owners who would rather their own theme handled presentation can switch that built-in design off in site settings, and Scaup publishes plain article HTML instead - For GitHub sites: creates actual files in the correct directory with proper frontmatter, following the structure of the site's existing pages. The customer can optionally pin a specific page for new pages to follow - For WordPress: creates a new post with a clean built-in design and a featured image picked from the site's media library, set as the post's real featured image so the theme, category pages and link previews all show it - New content is ALWAYS created for review before publishing - never auto-published without approval - Example: If competitors all have a "pricing" or "services" page and you don't, Scaup will create one targeting the relevant keywords ### Keyword Strategy and Opportunity Detection - Pulls real data from Google Search Console regularly (not estimated data from third-party tools) - **Country-scoped data**: Search Console pulls and keyword filters are scoped to the site's target country. A Bali hotel sees rankings for searches happening in Indonesia, not noise from people searching from other countries. This matters because GSC by default mixes every market together - a page can look like it's "ranking #4" globally while actually being #15 in the market that matters. - **Region-level focus**: For local businesses, we identify the specific area the business serves (city, region, metro area) and filter the plan and weekly report to keywords relevant to that area. A North-Israel interior designer doesn't get keyword recommendations for Jerusalem clients. This happens automatically - no setting to configure. - **Live SERP verification**: Keyword positions shown in reports are verified against live Google search results, not just GSC averages (which can be misleading because they blend positions across many queries and locations). The customer sees the actual rank a real searcher in their target country would see. - Identifies "striking distance" keywords: positions 8-20 where you're close to page one and a small content improvement could push you up - Spots keywords with high impressions but low click-through rate (your page shows up but people don't click). When this happens, Scaup automatically diagnoses the problem - is the title too generic? Is the page description missing? Is the content too thin? - and creates a fix. No manual intervention needed. - Detects climbing keywords that are gaining position week over week (worth doubling down on with dedicated content) - Finds keywords where you rank but have no dedicated page (opportunity to create one) - All of this feeds into the daily execution automatically. Up to 3 reactive fixes per week are created and applied alongside planned work. - The daily loop never sits idle: if the planned backlog runs low between monthly refreshes, Scaup pulls fresh improvement opportunities from Search Console so there's always something to work on that day. **Telling Scaup what to focus on (optional):** Site settings has two lists the owner can fill in: the searches they want to be found for, and pages they'd rather we left alone (a glossary, an author archive, anything they don't want touched). When the owner writes about their business in the "About your business" box, Scaup reads it and suggests both lists. The suggestions do nothing until the owner looks them over and turns them on, and they can be turned back off at any time without retyping anything. Leave both empty and Scaup works across the whole site, which is what happens by default. ### Competitor Intelligence - Automatically identifies the top competitors in your niche based on keyword overlap - Runs a gap analysis: keywords they rank for that you don't - Categorizes gaps by volume and difficulty to prioritize which ones to go after first - Refreshed monthly so the strategy stays current as your niche evolves - This is not just a report - the gaps directly feed into plan items that get executed ### Technical Files and Metadata - **robots.txt**: Controls which pages search engines can crawl. Scaup creates or updates it to ensure all important pages are crawlable. - **sitemap.xml**: A map of all pages on the site. Helps search engines discover and index everything. Scaup creates and maintains it. - **Structured data (JSON-LD)**: Machine-readable markup that helps Google understand what a page is about. Enables rich results in search (star ratings, FAQ dropdowns, event details, etc.). Scaup adds schema markup where appropriate. - **Image alt text**: Descriptive text for images. Helps with image search and accessibility. Scaup writes descriptive alt text for images that are missing it, in the site's own language. - **llms.txt**: A relatively new standard - a machine-readable summary of the site specifically for AI tools. Helps ChatGPT, Perplexity, Google AI Overviews, and other AI assistants understand and cite the site. Scaup creates and maintains it. ### Content Freshness - Google explicitly rewards sites that update regularly. A site that hasn't changed in 6 months signals "abandoned" to Google's algorithm. - Scaup's daily execution cycle means the site gets an improvement every single day - not just once a week. - This freshness signal compounds over time - Google learns to trust and re-crawl the site more frequently. Daily updates send a stronger signal than weekly batches. ### Internal Linking - When creating new pages, Scaup automatically adds links from existing relevant pages to the new page - When editing existing content, adds contextual links to other pages on the site where relevant - Safety checks prevent self-links, broken links, or linking to non-existent pages - For bigger structural linking changes (navigation menus, sidebars, footer links), Scaup flags opportunities with suggestions but doesn't auto-apply (these require code/template changes) --- ## Backlinks - The Complete Argument ### What backlinks are Backlinks are links from other websites pointing to yours. Google treats them as "votes of confidence" - if reputable sites link to you, Google trusts your content more. ### Why people think they're the most important thing Backlinks ARE the #1 ranking factor for competitive, high-volume keywords. "Best CRM software," "cheap flights to London," "project management tool" - these searches have 10+ well-funded companies fighting for position. In those battles, the site with more high-quality backlinks usually wins. ### Why that doesn't apply to most Scaup customers Most small sites and startups don't compete on those keywords. They win on long-tail, specific searches: - "yoga studio downtown Austin" - "invoice tool for freelancers" - "Astro portfolio template dark mode" - "emergency plumber Brooklyn open now" - "vegan meal prep service Portland" For these searches - which make up the vast majority of all Google queries - content relevance and quality are what decide the ranking. The page that best answers the specific question wins, regardless of backlink count. There simply aren't enough competitors with strong backlink profiles fighting for "vegan meal prep service Portland" for backlinks to be the deciding factor. ### The math A site with 50 pages, each targeting a specific long-tail keyword and answering the search intent perfectly, will drive more total traffic than a site with 5 generic pages and 200 backlinks. Long-tail keywords have less competition, higher intent (the searcher knows exactly what they want), and convert better. ### What Scaup does about backlinks Scaup doesn't build backlinks (no tool honestly can - it requires human action on external websites). But Scaup doesn't ignore them either: 1. **Niche analysis**: Identifies the most relevant places in the customer's niche to get listed 2. **Curated list**: Provides a specific, actionable list of best-fit websites: - Niche directories relevant to their business (e.g., "Best yoga studios in Austin" directories) - Community forums and discussion platforms where their audience hangs out - Industry listings and databases (e.g., Clutch for agencies, Product Hunt for SaaS) - Publications and blogs that accept guest contributions - Local business directories (Yelp, Google Business Profile, industry-specific) 3. **Not generic**: This is NOT a generic "go build backlinks" suggestion. It's specific places with direct relevance to their business, with context on how to get listed. The customer spends 30 minutes going through the list, not 30 hours researching where to submit. ### The compounding effect Better content naturally attracts more backlinks over time. People link to pages that are genuinely useful, comprehensive, and authoritative. The content work Scaup does today becomes the backlink magnet of tomorrow. This is a well-documented phenomenon: content quality drives organic backlink acquisition. ### How to handle the backlinks objection in sales Never dismiss the concern - the customer is right that backlinks matter. The argument is: 1. They're not the bottleneck for YOUR site at YOUR size and for YOUR keywords 2. Content quality is the thing you can control that makes the biggest difference right now 3. We still provide a curated backlink action list so you're not ignoring it 4. Our content improvements naturally attract backlinks over time That's the complete answer. Don't get defensive. Acknowledge, reframe, then show the full picture. --- ## What Scaup Guides You Through ### Content Requiring Manual Review Not everything can be safely auto-applied: - Homepage rewrites (too high-stakes for auto-apply) - Major content overhauls on important pages - Changes that our validation system flags as risky - Content on page-builder pages (Elementor, etc.) that can't be safely edited externally For these cases, Scaup drafts the improvement and presents it for review with a clear before/after preview. The customer approves or rejects, and the approved changes are applied. ### AI Visibility Tracking - Scaup measures how often AI tools (ChatGPT, Google AI) mention the site vs its competitors - Shows which sources AI assistants pull answers from in the site's niche, and which of the customer's own pages AI tools cite most - Tracks how the customer's AI mentions trend month over month, using real historical data - Surfaces specific questions where AI cites a competitor but not the customer, then feeds those gaps directly into the growth plan as content to create - Scaup also asks ChatGPT, Claude, Gemini, and Perplexity the customer's most important questions directly, then records whether each answer names or links to the site. That is what a real searcher sees, not an estimate - Every month, Scaup re-asks the questions it has been working on and checks whether AI assistants have started recommending the site for them. When they have, those wins appear on the dashboard and in the monthly email ("AI assistants now recommend you for 2 more questions than last month"). Only real flips, never a made-up score - The customer can see how often AI tools mention them, how it trends over time, and what Scaup is working on to improve it - This runs automatically every month. No setup required. --- ## Supported Platforms - Detailed ### GitHub-Connected Sites **Supported frameworks:** Astro, Next.js, Gatsby, Nuxt, SvelteKit, Remix, Vite + React (which covers Lovable, Bolt and v0 sites), and static HTML sites. **How it works:** Scaup clones the repository, analyzes the file structure, identifies editable content areas (page titles, headings, paragraphs, meta tags, frontmatter fields, markdown content), makes targeted edits, validates the changes, runs the build to verify nothing breaks, and commits. **Astro gets the deepest support:** - Full recipe system for creating new pages with correct file structure - Component registration (adding imports and usage) - Structured data insertion in page heads - Navigation link additions - Content file creation (.md, .mdx, .txt, .json, .xml) **All other frameworks get:** - Content editing (titles, headings, paragraphs, meta tags, frontmatter, markdown) - New page creation (using existing pages as templates, or a customer-pinned page if they've set one) - Technical file creation/updates (robots.txt, sitemap.xml, llms.txt) **What frameworks Scaup does NOT support:** - Custom or proprietary frameworks - Sites that don't output standard file-based content - If the framework outputs standard HTML files, there's a chance it works under static HTML support ### Lovable (and Bolt, v0) **Supported:** Sites built with Lovable, Bolt or v0, whose code lives in a GitHub repository. **How it connects:** Lovable has its own tile in setup, but the connection is the GitHub one. The customer picks their repository and Scaup works out the rest on its own. Nothing to configure and nothing to install. **What Scaup can do on a Lovable site:** - Improve the wording on existing pages - Write and publish new pages, wired into the site's navigation so visitors and Google can reach them - Create and maintain robots.txt and sitemap.xml - Improve the home page title and description **Per-page titles:** A site built this way starts out sharing one title and one description across every page, which means Google shows the same line for the pricing page as for the about page. Scaup can fix that, but it means a small one-time change to the customer's own project, so we ask during setup instead of doing it quietly. If they say yes, Scaup makes that change when it can: it adds one library and gives each page its own title and description, starting from the ones the site already has, then improves them one page at a time. Some projects are set up in a way we will not touch without risking their build, and on those we leave the project alone and tell them. Until that one-time change has run, titles and descriptions stay as they are, the home page included. Everything else carries on either way, and they can change their mind later from site settings. **Publishing:** If the site is hosted by Lovable itself, the live site only updates when someone clicks Publish inside Lovable. There is no way for Scaup, or anyone else, to click it from outside. So Scaup does the work, then tells the customer in the dashboard and in the weekly email that changes are ready and waiting. If nothing has been published for a week, Scaup pauses new work rather than letting it stack up unseen, and starts again as soon as they publish. Customers who would rather never think about that step can connect their Lovable project to Netlify or Vercel and point their domain at it. Both rebuild the site on their own, and Scaup goes back to being completely hands-off. Scaup works out which of the two situations a site is in by itself. ### WordPress **Supported:** Standard WordPress.com and self-hosted WordPress installations. **How it connects:** Through WordPress Application Passwords, a built-in authentication feature since WordPress 5.6 (December 2020). The customer creates a dedicated Application Password for Scaup in their WordPress admin panel. Scaup uses the standard WordPress REST API. No FTP access, no admin password sharing. **SEO plugin support:** Yoast SEO, Rank Math, and All in One SEO (AIOSEO). These plugins expose an API that lets Scaup write SEO titles and meta descriptions. Without a SEO plugin, Scaup can still update titles, post content, create posts, and manage technical files - but meta descriptions require one of these plugins. Rank Math free is the easiest to install (2 minutes) and is usually recommended during onboarding. **Scaup WordPress plugin:** A lightweight optional plugin that adds API endpoints for advanced features: automated meta description writes and static file management (robots.txt, llms.txt). The plugin is open-source and does nothing else. **New page design:** New WordPress pages use a clean, built-in design that renders well on any theme or page builder and inherits the site's fonts, so there's nothing to configure. On GitHub sites, the customer can optionally pin a page from site settings for new pages to follow its structure. ### Wix **Supported:** Wix websites with the Wix Blog app installed. **How it connects:** Through Wix's official OAuth 2.0 app installation. The customer clicks a link to install the Scaup app on their Wix site, approves the permissions, and they're connected. No passwords to enter, no plugins to install manually. **What Scaup can do on Wix:** - Create and publish blog posts with full SEO fields (title, description, slug, social sharing tags), formatted with real headings, bullet lists, and bold text in the site's own blog style (not one wall of text) - Update existing blog post content and SEO fields - Manage blog categories and tags - Update product descriptions and SEO fields (if Wix Stores is installed) - Update CMS collection items (if Wix CMS is used) - Update robots.txt **What Scaup can't do on Wix:** - Edit static page content (Home, About, Contact, etc.) - Wix doesn't expose these pages through their API - Create or manage URL redirects - Update image alt text **New content creation on Wix:** Blog posts are created and published. Starter plan includes up to 4 new posts/month, Pro up to 8. This matches the same organic growth approach used on other platforms. **Major app updates on Wix:** Wix applies minor Scaup updates automatically, but major versions need the site owner to tap Update in their own Wix dashboard. When a major update is waiting, Scaup shows a banner on the site's dashboard with a direct link to the right place in Wix. Work continues as normal while the update is pending; the banner disappears automatically once the owner approves, or immediately when they tap "Check again." ### Webflow **Supported:** Webflow websites, connected via OAuth. **How it connects:** Through Webflow's official OAuth 2.0 flow. The customer clicks "Connect Webflow," logs in, and approves Scaup's permissions. No passwords to enter, no plugins to install. **What Scaup can do on Webflow:** - Update SEO title and meta description on static pages (not just blog/CMS content - a real advantage over Wix, which can only touch blog posts) - Create and publish new CMS collection items with SEO fields - Update existing CMS collection item content and metadata **What Scaup can't do on Webflow:** - Edit static page body content (Webflow's visual Designer owns that, no API for it) - Create or manage URL redirects - Update image alt text - Customize sitemaps **New content creation on Webflow:** New CMS collection items publish automatically as part of the same daily, paced cadence used on every other platform. **Publishing and the customer's own unpublished work:** Webflow publishes an entire site at once, so a publish would also push whatever the owner has left unfinished in the Designer. Scaup checks for unpublished Designer edits before it applies anything. If it finds any, it makes its changes but holds back the publish, and the site page shows a note explaining why. Scaup's work then goes live alongside the owner's next publish. Nothing of the customer's is ever published without them choosing to. ### WordPress Limitations - Be Honest About These **Page builders (Elementor, Divi, WPBakery, Beaver Builder, Bricks, Oxygen):** Page builders store content in their own proprietary format - encoded blocks, shortcodes, nested JSON. Editing that content from outside the page builder is dangerous: one wrong character and the entire page layout breaks. Scaup detects page builder content and deliberately skips those pages. **What Scaup can still do on page-builder sites:** - Page titles (not stored by the page builder, so fully editable) - Meta descriptions via Yoast, Rank Math, or All in One SEO (also outside the page builder) - New blog posts using the native WordPress editor (not the page builder) - Technical files (robots.txt, sitemap.xml, llms.txt) - Competitor analysis and keyword strategy (doesn't depend on how the site is built) - Backlink suggestions (niche analysis is platform-independent) **For page-builder content itself:** Scaup drafts improvement suggestions and delivers them as downloadable files. The customer pastes the improved content into their page builder manually. They still get the strategy and the writing - they just apply it themselves. **Managed hosting (WP Engine, Kinsta, Flywheel):** Some managed WordPress hosts disable Application Passwords as a security measure. Scaup detects this automatically and shows a clear error with workaround instructions. It's fixable but should be mentioned upfront if the customer uses one of these hosts. **Sites behind firewalls (Cloudflare, etc.):** Scaup's worker connects from a fixed set of IP addresses. If the site is behind Cloudflare or another firewall, the customer adds those IPs to their allowlist. Scaup shows the exact IPs in site settings. One-time setup, takes 2 minutes. --- ## Safety and Control - Detailed ### Three Layers of Validation Every change goes through three safety gates before reaching the customer's site: 1. **Code validation**: The edited file is parsed to verify it's still valid syntax. For GitHub sites, this means the file structure (HTML, markdown, frontmatter, etc.) is intact. If the edit would create invalid syntax, the change is rejected. 2. **AI content review**: A separate AI model reviews the applied changes and checks for: - Accidental removal of important business information (phone numbers, addresses, key facts) - Significant content shrinkage (removing more than it adds) - Loss of keywords the page was already ranking for - Coherence and readability issues 3. **Build check (GitHub sites only)**: Scaup runs the site's actual build command (npm run build) and verifies it passes. If the change would break the build, it's rejected before it reaches the repository. The customer's CI/CD pipeline never sees a broken commit. If ANY layer fails, the change is rejected. It never reaches the site. ### Change Stability - We Don't Churn the Same Page When Scaup sets a page's title or description, it leaves that change in place to settle. Search engines need time to trust a new title and see how people respond to it, so rewriting the same title repeatedly works against the customer. After Scaup tunes a page's title/meta, it holds that page for about a month before it would consider revisiting it - daily work moves on to other pages and other kinds of improvements in the meantime. The result is steady progress across the whole site rather than the same page being changed every week. ### User Control **Automatic (the default, per category):** - Changes go live as Scaup makes them, and the customer gets an email saying what changed - Set separately for new pages, content edits and technical fixes - Everything stays reversible **Review mode (opt-in, per category):** - Every proposed change is shown as a before/after diff - The customer sees the exact current text and the exact proposed replacement - They approve or reject each change individually - For an image description, the customer can also edit it directly before approving, not just accept or skip it - Nothing goes live without explicit approval - This is the default because trust needs to be earned **Auto-apply (opt-in):** - A customer who wants a gate turns review on for whichever categories they care about - Left alone, changes go live immediately and stay one-click reversible for the current week **Reversibility:** - GitHub sites: every change is a git commit with a stored SHA. Revert is a standard git revert - one click. - WordPress: content snapshots are stored before every change. Restore is one click. - There is NO scenario where a change is permanent and the customer is stuck. **New content handling:** - New pages and blog posts are ALWAYS created as drafts, never auto-published - The customer reviews in their CMS/editor and publishes when ready - This is true even in auto-apply mode ### Content-Only Edits - Hard Boundary Scaup NEVER touches: - Code (JavaScript, TypeScript, Python, etc.) - Design or layout (CSS, Tailwind classes, component structure) - Components (React, Vue, Svelte components) - Imports, exports, or dependencies - Configuration files - Images or media files (except writing alt text) Scaup ONLY edits: - Text content (paragraphs, list items, blockquotes) - Page titles and meta descriptions - Headings (H1-H6 text, not hierarchy) - Frontmatter string fields (title, description, etc.) - Markdown body content - HTML text nodes - Meta tag attributes (description, alt text) This boundary is enforced at the system level. The system literally cannot output code, component tags, CSS, or imports. It's not a guideline that might be violated - it's a hard technical constraint. ### Anti-AI Writing - Quality Control All content goes through filters that detect and strip AI writing patterns: **Banned patterns (50+):** - Filler phrases: "furthermore," "moreover," "it's important to note," "in conclusion" - AI-telltale vocabulary: "comprehensive," "leverage," "seamless," "personalized," "robust" - Formatting tells: em dashes, excessive bullet points, "Rule of Three" constructions - Generic openings: "In today's digital landscape," "In the world of," "When it comes to" - Inflated language: "revolutionize," "transform," "game-changing," "cutting-edge" **Content integrity rules:** - Never fabricates facts, statistics, awards, certifications, or years of experience - Never creates fake testimonials, reviews, or customer quotes - Never invents company history or claims - Only uses information that exists on the site or is publicly verifiable - If a page says "10 years of experience," Scaup keeps that. It never changes it to "15 years" to sound more impressive. The result: content that reads like a human copywriter wrote it, because we actively prevent it from reading like a machine did. --- ## Ranking Protection A hacked site can drop out of Google overnight, and most owners don't find out until their traffic collapses. Ranking Protection watches every connected site for the security problems that specifically hurt search rankings. It is live in production and runs across every platform. On the customer's dashboard this appears as the "Ranking protection" card. **What it checks (every platform: WordPress, Wix, Webflow, GitHub-hosted):** - **Google blacklist status:** whether Google has flagged the site as unsafe for malware, phishing, or unwanted software. This is the single biggest warning sign that a site is about to disappear from search results. - **Injected spam links:** hidden links pointing to gambling, pharma, or other junk, slipped into the site by an attacker to boost their own sites using yours. We check the live pages and the site's saved settings (a common hiding spot), and when we find hidden links we show the owner the exact addresses that were planted, so they can see precisely what to remove. - **Injected malicious code:** hidden scripts and redirects buried in the pages or the site's saved settings. We read the page a visitor actually gets and flag what doesn't belong. - **Secure-connection check:** whether the site loads securely over https, shown by the padlock in the browser. Google favors secure sites and visitors leave when a browser shows a "Not secure" warning, so we check the site loads securely and flag it when it doesn't. - **Spam outbound links:** whether the pages link out to sites known for spam. Google judges a site partly by who it links to, so links to a bad neighborhood drag rankings down. We flag it only when several such links show up, so a single stray link can't trip a false alarm, and we list the exact sites. - **Cloaking check:** whether a page shows a search engine something different from what real visitors see. That trick, called cloaking, is a classic sign of a hacked site and Google penalizes it heavily. We compare the two views and flag only a genuine mismatch, so normal things like showing ads or a logged-in view can't trip it. **Deeper scan on WordPress (widest exposure):** - **Nameless dropper add-ons:** when a WordPress site is broken into, attackers often leave behind a fake add-on with a harmless-looking name like "WordPress Fix" and no real maker behind it. Legitimate add-ons always declare who made them and what they do. We flag the blank, anonymous ones. - **Elementor page layouts:** on sites built with Elementor, spam links and injected code can hide inside the page-builder layout, where a normal content scan doesn't look. We read the Elementor layout too and flag anything that doesn't belong. **What it does NOT do:** - It never changes, updates, or deletes anything. Detection only. A wrong automatic change to a hacked live site can do more damage than the hack, so when we find a problem we tell the owner what it is and what to do next. - It is not full antivirus. It covers the security problems that hurt Google rankings, not every possible threat. - It does not chase outdated-plugin version numbers. Those rarely affect rankings on their own, so the product focuses on what actually harms search visibility. **Cleanup as a paid service:** when a scan finds a real infection, Scaup can handle the hands-on cleanup as a separate paid service, only with the owner's go-ahead. This is a natural upsell. **Runs on its own every week:** every connected site is checked once a week automatically, with no one needing to press a button. If a check turns up something serious, the owner gets a single email in plain language that says what we found and what to do about it. We only email when a new problem appears, so an owner never gets the same warning week after week. The email is deliberately honest that some of these are things the owner or their web person needs to clean up, not something Scaup changes for them. This runs for every paying site on every plan. **Where it stands:** the detection engine, the automatic weekly scan, the owner alert email, and the customer-facing safety card on the dashboard are all live. The card shows a calm "nothing is hurting your ranking" summary with each check (Google blacklist, hidden junk, suspicious add-ons, secure loading, spam outbound links, cloaking) in plain language, and every check has a small info icon that explains what it looks at in one non-technical sentence. The calm "your site is protected" line in the recap email is rolling out next. **How to talk about it:** "Set it up once and we make sure nobody quietly wrecks your rankings." This is the one feature where Scaup drops its always-positive tone on purpose. A green "all clear" that hid a real infection would destroy trust, so when something is wrong we say so, clearly and specifically. --- ## Security & Data Protection ### Where Data Is Stored - **Database and backend:** All application data (sites, plans, reports, settings) is stored on Convex, a cloud-hosted backend platform running on AWS infrastructure. - **Frontend hosting:** The web app is hosted on Netlify as a static site. No user data is stored on the frontend. - **Background jobs:** Daily site improvements run on a Railway-hosted worker. The worker is stateless and doesn't store data locally. ### How Credentials Are Protected Every credential Scaup stores is encrypted with **AES-256-GCM** before it reaches the database. Each value gets its own random initialization vector (IV), so identical passwords produce different encrypted values. The encryption key is stored as an environment variable, separate from the database. | Credential | Purpose | Storage | |---|---|---| | Google OAuth tokens | Read Search Console data, submit pages to Google index | AES-256-GCM encrypted | | WordPress app password | Push content changes to WordPress sites | AES-256-GCM encrypted | | WooCommerce API keys | Access product data on WooCommerce stores | AES-256-GCM encrypted | | Wix OAuth tokens | Push content changes to Wix sites | AES-256-GCM encrypted | | GitHub connection | Push content changes to repositories | OAuth via GitHub App (no token stored) | Google OAuth access tokens expire after one hour and are refreshed automatically using the encrypted refresh token. ### Data Isolation - Every backend request re-verifies user identity from the authentication session. User IDs passed from the browser are never trusted. - Every query and mutation validates resource ownership before returning data or making changes. - Database queries always filter by authenticated user first. No API endpoint returns data across multiple users. - Collaborators get limited access. They can't change billing, delete sites, or view stored credentials. ### Deleting Your Data - Removing a site deletes everything stored for it: the growth plan, reports, keyword history and page records. The owner can do this from Site Settings at any time. - Some platforms never tell us when the app has been removed. We find out only because our requests start being refused, and a refusal can also be a temporary fault. So we delete the stored access credentials straight away, stop calling the platform, and keep the work itself, because reconnecting has to bring the customer back to what they had. - When that happens, the notice on the site offers a direct route to delete it all instead. The choice is the customer's, and we never delete on a guess. - We keep the seven most recent daily backups, so a deleted copy stays inside those until newer ones displace it. ### Authentication Scaup uses Google OAuth for sign-in. We never see or store Google passwords. Google handles the auth flow and we receive an identity token confirming who the user is. ### What We Access On Connected Sites - **Google Search Console:** Read search queries, clicks, impressions, page performance, indexing status. Submit URLs for re-indexing. Cannot delete properties or change settings. - **WordPress:** Read and update posts, pages, and metadata. Cannot install plugins, change themes, or modify settings. - **GitHub:** Read files, create branches, open pull requests. Cannot merge to main without approval, access other repos, or modify repo settings. ### AI and Customer Data Scaup uses Anthropic's Claude API for content analysis and writing. What gets sent: page content, search performance data, competitor analysis. What doesn't get sent: credentials, login tokens, personal information, other users' data. Per Anthropic's API terms, data sent through the API is not used to train models and is not retained beyond request processing. ### Worker and API Security - Worker-to-backend communication uses a shared secret verified with timing-safe comparison. - All traffic between systems uses HTTPS/TLS encryption in transit. - All API keys and encryption keys are stored as environment variables in each platform's secret management. Nothing hardcoded. ### Honest Gaps - No SOC 2 or compliance certifications yet. Infrastructure providers (AWS, Convex, Railway, Netlify) maintain their own certifications. - Scaup is an early-stage product. Security practices are solid but formal audits haven't happened yet. --- ## What Scaup Does NOT Do - And Why It Doesn't Matter ### Things the customer's hosting/framework already handles These are real ranking factors, but they're infrastructure-level, not content-level. Modern hosting and frameworks handle them automatically: - **Page speed** (image compression, lazy loading, CDN, caching): Vercel, Netlify, Cloudflare, and most modern hosts handle this. AI-built sites are typically fast by default. If there's a genuine speed issue, it's a one-time fix (a few hours of developer time), not an ongoing service. Google has confirmed page speed is a tiebreaker, not a primary ranking factor - it matters when two pages are equally relevant, but content relevance still decides 95% of rankings. - **Mobile responsiveness**: Every modern framework and template is mobile-first. This hasn't been a real issue since 2018. - **HTTPS/SSL**: Free and automatic on every major host since 2020. - **URL redirects (301/302)**: Framework-level configuration. Scaup doesn't rename existing URLs because that would break existing links and require redirect setup. - **Image file optimization** (WebP, AVIF, compression): Build tools and CDNs handle this automatically. Scaup handles alt text (the SEO-relevant part) but not file optimization. ### Things that require structural changes - **Site architecture and navigation**: Menus, breadcrumbs, category pages, footer links. These are design/code decisions that affect the entire site structure. Scaup works within the existing structure and suggests improvements but can't modify layout templates. - **Heading hierarchy**: Scaup can edit heading TEXT but can't add, remove, or reorder heading LEVELS (e.g., convert an H3 to an H2). That's a structural code change. - **JavaScript rendering / SPA SEO**: If a site is a single-page app that renders content client-side, search engines may not see the content. Fixing this requires SSR/SSG setup - an architectural change, not a content change. ### Things outside the product scope - **Google Business Profile / Local SEO**: Google Maps listing, NAP (Name/Address/Phone) consistency, review management. This is a separate platform and a separate discipline. - **Social media management**: Not related to search ranking. - **Paid advertising** (Google Ads, Meta Ads): Pay-per-click, not organic. - **Email marketing**: Different channel entirely. - **Domain authority**: Can't be directly changed by any tool. It builds over time through content quality, backlinks, and consistent publication. Everything Scaup does contributes to this indirectly, but there's no "domain authority boost" button. --- ## SEO Weight Distribution - What Actually Matters Not all SEO factors are equal. Here's how they rank by actual impact on search rankings, with Scaup's coverage: **Factors Scaup fully automates (~70% of actionable SEO weight):** - Content quality and relevance: 9.5/10 impact - Keyword targeting: 9.0/10 - Page titles: 8.5/10 - New content creation: 8.5/10 - Competitor gap analysis: 7.0/10 - Headings: 6.0/10 - Search Console monitoring: 6.0/10 - Content freshness: 5.5/10 - Structured data: 5.0/10 - Meta descriptions: 4.0/10 - Sitemap + robots.txt: 3.8/10 - Image alt text: 3.0/10 - AI visibility tracking: 6.5/10 - llms.txt: 2.0/10 **Factors Scaup guides you through (~15% of SEO weight):** - Backlinks: 9.5/10 impact (we provide the action list, you submit) - Internal linking: 8.0/10 (partially automated, suggestions for structural changes) **Factors handled by your stack/hosting (~15% of SEO weight):** - Page speed: 7.0/10 - Mobile responsiveness: 6.5/10 - HTTPS: 3.5/10 - Redirects: 4.0/10 --- ## How Scaup Compares - Detailed ### vs Semrush / Ahrefs ($130-500/month) **What they do:** Keyword research tools, site audits, rank tracking, backlink analysis. They generate reports, dashboards, scores, and lists of "issues." The customer is expected to understand the reports and implement everything themselves. **The gap:** 90% of Semrush/Ahrefs users never act on the data. They see 200 issues, feel overwhelmed, and close the tab. The tool is built for SEO professionals, not founders. **Scaup's angle:** "They tell you what's wrong. We fix it. Every day. You don't even need to log in." ### vs Yoast SEO / Rank Math (free - $100/year) **What they do:** WordPress plugins that show a per-page SEO checklist inside the editor. "Your title is too short." "You haven't used the focus keyword in the first paragraph." Manual, one page at a time, reactive (only when you're editing a page). **The gap:** Requires the customer to manually optimize every page, every time they edit. No data-driven strategy, no competitor analysis, no automated execution. **Scaup's angle:** "Yoast tells you what to fix on one page. Scaup fixes all your pages, every day, based on actual Google data." ### vs Freelance SEO ($500-2000/month) **What they deliver:** Typically a monthly PDF report, a handful of title/description rewrites, maybe a blog post, and a monthly call explaining what you should do next. Between check-ins (3-4 weeks), nothing happens. **The gap:** Implementation gap. Most freelancer recommendations require the customer or a developer to actually implement them. And monthly cadence is too slow - SEO is a daily game. **Where freelancers genuinely win:** Creative strategy, brand voice judgment, link-building relationships, PR outreach. These are human-only skills. **Scaup's angle:** "80% of what a freelancer bills for is on-page content work. We do that faster, cheaper, and 30x more frequently - daily vs monthly. The other 20% (relationships, creative strategy) is where humans still win. The question is whether that 20% is worth $500+/month right now." ### vs Surfer SEO ($99-199/month) **What they do:** Content optimization scoring. You write content, Surfer scores it and suggests keyword additions. Manual workflow - you still write, edit, and publish yourself. **The gap:** Another tool that tells you what to do but doesn't do it. Also doesn't handle technical files, competitor analysis, or automated execution. **Scaup's angle:** "Surfer scores your content. Scaup writes and applies it." ### vs ChatGPT (free - $20/month) **What it does:** Writes content when you ask it to. **The gap:** ChatGPT does ~10% of the SEO job. The other 90% is: - Knowing WHAT to write (which keywords, which pages, which gaps) - requires GSC data + competitor analysis that ChatGPT doesn't have - Safely APPLYING it (editing a live codebase without breaking things) - Doing it CONSISTENTLY every day (you'll use ChatGPT once, then get busy) - Not BREAKING things (preserving OG tags, canonical URLs, structured data, existing rankings) - STRATEGY (ChatGPT doesn't tell you that your homepage title is wrong or that you're missing 12 keyword opportunities) **Scaup's angle:** "ChatGPT is a pen. Scaup is the whole marketing department." ### vs Doing Nothing **The cost:** Invisible but real. Google rewards sites that update regularly. Competitors who optimize daily will outrank you over time. Every day of inaction is a day where someone else is taking the search traffic you could have had. **Scaup's angle:** "You didn't start your company to do SEO. But your competitors' sites are getting optimized every day. The question isn't whether to do this - it's whether to do it yourself or let us handle it." --- ## Timeline Expectations - Be Honest **Week 1:** Growth plan generated, first batch of improvements applied. Customer can see exactly what changed. **Weeks 2-4:** Daily improvements continue. Pages start getting indexed by Google. First impressions may appear in GSC for new content. **Month 2-3:** Meaningful traffic starts for new sites. Existing sites with some authority see faster results. Keyword movements become visible in weekly reports. **Month 3-6:** Compounding effect kicks in. More pages ranking = more keywords = more traffic. Content freshness signals compound. New content starts attracting organic backlinks. **Ongoing:** Daily improvements prevent stagnation. Monthly competitor refreshes keep the strategy current. The site continuously gets better, not just once. **Important:** Google deliberately makes ranking a slow, trust-based process to prevent manipulation. Anyone promising first-page results in 2 weeks is either lying or doing something that will get the site penalized. Set realistic expectations early - it builds trust. --- ## Pricing Two plans, priced per site, per month. Everything is included in both except where noted. **Starter - $149/month** - Content optimization and daily execution - New page creation (2 new pages per month) - Competitor analysis and monthly refresh - Technical file management (robots.txt, sitemap, structured data, alt text) - Backlink suggestions - Google Search Console monitoring - llms.txt for AI crawlers - 1 team member **Pro - $249/month** - Everything in Starter, plus: - AI visibility tracking (who AI cites, competitor gaps, content to fix it) - 6 new pages per month - 3 team members No per-page charges, no per-keyword charges, no hidden AI usage fees, no setup fees, no annual commitment required. **Comparison math for the customer:** - Freelance SEO: $500-2000/month, works on your site once a month - Semrush: $130/month minimum, you do all the work yourself - Content writer: $100-500 per article, no strategy or implementation - Scaup: from $149/month for all of the above, every day --- ## Cancellation and Risk - Cancel anytime, no penalties, no lock-in period - All changes already made stay on the site permanently - Nothing gets removed, reverted, or locked down - The customer keeps all optimized content, new pages, and technical files forever - Even 2 months of Scaup leaves permanent improvements **The low-risk pitch:** "Connect your site, see the growth plan we generate. If you don't like it, disconnect. You've spent 5 minutes. If you try it for a month and it's not for you, cancel and keep everything we built." --- ## Common Objections - Quick Reference **"It's too expensive."** Compare to alternatives: freelancer ($500-2K/mo), Semrush ($130/mo + your time), content writer ($100-500/article). Calculate the value of their time: if they bill $100/hr and SEO takes 5 hours/month, that's $500 in time alone. **"I don't trust AI to edit my site."** Turn on review and nothing goes live without approval. 3-layer validation. Content-only (never touches code). Everything reversible. "Switch review on, judge the first batch. If you don't like it, you've lost nothing." **"Backlinks are more important."** See the full Backlinks section above. Don't dismiss - acknowledge, reframe to long-tail keywords, mention the curated list, explain the compounding content effect. **"Google penalizes AI content."** Google's policy (reaffirmed in the March 2024 spam update) targets "scaled content abuse" (mass-produced thin or near-duplicate pages), not AI as a method. We're built so we can't accidentally produce that pattern: every new page has to pass a substance gate (real user question + real facts + a reason it's better than what already ranks), and if it doesn't, we skip rather than publish filler. We hard-cap how many related pages can exist per topic, so we can never produce doorway-style location-swap or product-swap variants. On top of that, anti-AI-writing filters strip the obvious machine-writing patterns from every output. The question Google asks isn't "AI or human." It's "helpful or not." **"I'll just use ChatGPT."** Writing is 10% of the job. Strategy, safe application, daily consistency, and not breaking things are the other 90%. "Will you actually do this every day? Pull GSC data, analyze competitors, edit your codebase, validate the build?" **"Can you guarantee rankings?"** No, and anyone who does is lying. Google uses 200+ signals. What we guarantee is the work gets done, every day, based on real data. Sites that do the right things consistently almost always improve over time. **"SEO is dead, it's all AI now."** Google still drives 90%+ of search traffic. AI search is additive. AI tools cite the same well-structured content that ranks on Google. We optimize for both simultaneously. **"My site uses Elementor/Divi."** Be honest: we can't safely auto-edit page builder content. But we CAN optimize titles, meta descriptions, create blog posts, manage technical files, do competitor analysis, and provide backlink suggestions. For page builder content, we draft suggestions you apply manually. Set expectations clearly. **"How is this different from a freelancer?"** Weekly vs monthly. Automated vs manual. Data-driven vs opinion-based. 80% of freelancer work is content optimization we automate. Honest trade-off: freelancers bring human creativity and link-building relationships. But for the on-page work that makes up most of their billable hours, we're faster, cheaper, and more consistent. --- ## Key Talking Points 1. "Your AI built the site. Ours gets it found on Google." 2. "We're not an SEO tool that gives you reports. We're the team that does the work." 3. "Connect once, we handle everything. We improve one thing every day and send you a weekly recap." 4. "Backlinks matter for huge competitive keywords. For your size, content quality is what decides the ranking. We handle the content - and give you a list of where to get backlinks." 5. "Every change is reversible, and if you turn on review nothing goes live without your approval." 6. "We don't guarantee rankings - anyone who does is lying. We guarantee the work gets done, every day, based on real data." 7. "It takes less time to set up Scaup than to read a single Semrush audit." 8. "Even if you cancel after a month, you keep everything we built. There's no downside." 9. "80% of what an SEO freelancer does, we do automatically. Every day, not once a month." 10. "Google rewards consistency. We're the only tool that actually shows up and does the work every single day."