01 What this engine does
The Elite SEO Engine is your daily research workbench. Instead of buying a $99/month seat at SEMrush or Ahrefs, you get a focused 5-tool toolkit that does the work that actually moves rankings: find keywords, decode the SERP, reverse-engineer competitors, audit your own pages, and brief your writer (or AI) before they touch a keyboard.
The engine pulls live data from Google Suggest + live SERPs + real page HTML, then Claude reads everything and writes the strategy. No keyword volume guessing — just intent buckets, real top-10 results, and decisions you can act on.
02 The 5 tools at a glance
| Tool | Use when… | Time | Cost |
|---|---|---|---|
| Keyword Research | You're starting fresh in a niche and need 50 ideas + clusters. | 10–20s | 3 cr |
| SERP X-Ray | You picked a keyword — now you need to know what the SERP actually rewards. | 10–15s | 4 cr |
| Competitor Snap | One specific page is outranking you and you need to know why. | 12–18s | 4 cr |
| Page Optimizer | You have a live page that should be ranking but isn't. | 10–15s | 3 cr |
| Content Brief | You're about to write — and you want the piece to rank top-5. | 15–25s | 6 cr |
03 The full workflow
You don't need every tool every time, but this is how they fit together when you're starting a new piece of content from scratch:
04 Tool 1 — Keyword Research
Drop a seed keyword. We hit Google Suggest with 12 modifier prefixes (how to, best, vs, review…) to expand the seed, then Claude curates the raw list into a usable keyword set with intent and difficulty buckets.
Inputs
- Seed keyword (required) — 2–4 words is the sweet spot
- Location (optional) — US / UK / India / Canada / Australia / Global
Outputs
- 1–2 sentence niche overview
- Top 5 best opportunities (low difficulty + commercial intent)
- 3–6 topical clusters (one cluster = one piece of content)
- Up to 50 keywords with volume / difficulty / intent / type
- One-click CSV export
Pro tips
- Don't go too narrow. A seed like
"red shoes for women"returns 8 keywords. Try"red shoes"instead — broader seed, richer cluster output. - Trust the buckets, not exact volumes. We use
high / medium / low / tinyinstead of fake 12,400/mo numbers because nobody knows the real volume — Google has stopped publishing it for over a decade. - Mine the clusters, not the table. Each cluster is a content piece. Twenty clusters = twenty months of content calendar.
- Filter by intent. If you're monetizing, the table's
commercial+transactionalrows are where the money is.informationalrows are top-of-funnel.
05 Tool 2 — SERP X-Ray
Paste a target keyword. We pull the live top-10 organic results (via DuckDuckGo's HTML index, which mirrors Google closely without API limits) and Claude reads all 10 to figure out: what format dominates, what's the intent, what every page must cover, and where the gaps are.
Inputs
- Target keyword (required) — the exact phrase you want to rank for
Outputs
- Dominant format (listicle / review / guide / video…)
- Search intent classification
- Must-cover topics (table stakes for ranking)
- Content gaps (your differentiation lever)
- 3 ranking angles ranked by ranking probability
- Recommended target word count
- Live top-10 list with snippets
Pro tips
- Run X-Ray before writing. The biggest mistake new SEOs make is writing a 1,500-word blog post for a SERP that's ranking 5-minute videos. Match the format or lose.
- Take the gaps seriously. If Claude flags
"none of the top 10 includes pricing comparisons", that's a free win — add the comparison table. - The target word count is a floor, not a ceiling. Aim for the recommended length + 15%.
06 Tool 3 — Competitor Snapshot
Paste any public URL. We fetch the page, parse the HTML (title, H1–H3, body, schema, link patterns, image coverage), and Claude reverse-engineers what the page is really doing — so you can copy the parts that work and exploit the parts that don't.
Inputs
- Page URL (required) — full URL including https://
Outputs
- Best-guess primary keyword
- 5–10 secondary keywords being woven in
- Content angle / unique POV
- Topical clusters covered
- Word count, schema, link counts, image alt coverage
- Monetization mechanism (ads / affiliates / opt-in / product)
- 3–5 exploitable weaknesses
- 2–3 sentence outrank strategy
Pro tips
- Pick the actual ranking page, not the homepage. If they rank #1 for
best ai tools 2025, snap/best-ai-tools-2025/, notexample.com. - Snap the top 3 results. Three competitor snaps + one SERP X-Ray = a complete picture for ~16 credits.
- JS-heavy pages return thin data. Many Webflow / Notion / SPA pages render content via JavaScript that our scraper doesn't execute. If word count comes back at
200on a page that's clearly longer, that's why.
07 Tool 4 — Page Optimizer
Paste your own page + the keyword you want it to rank for. We score 0–100 across four pillars and hand you a prioritized fix list. This is the one you run after you've published a piece and it's not ranking.
Inputs
- Your page URL (required) — must be live + public, not localhost or staging
- Target keyword (required) — the keyword this specific page should rank for
Outputs
- Overall score 0–100 + verdict (Weak/Decent/Strong/Excellent)
- Sub-scores for Title / Meta / Content / Technical
- Wins list (what's already working — don't break it)
- Prioritized fix list (critical → high → medium → low)
- Rewrite suggestions for title + meta description
- Raw page diagnostics (word count, links, alt coverage, schema)
How to read the score
| Score | Verdict | What it means |
|---|---|---|
| 85–100 | Excellent | You're in great shape. Watch for ranking, don't over-optimize. |
| 70–84 | Strong | Solid foundation. Fix the criticals + highs and you'll likely climb. |
| 50–69 | Decent | Real fixable issues. Work the prioritized list top-down. |
| 0–49 | Weak | Likely has a fundamental issue (wrong intent, off-target keyword, thin content). |
08 Tool 5 — Content Brief
The highest-leverage tool in the engine. Pick a target keyword + page type. We pull the live SERP and Claude builds a complete writer's brief — title options, meta description, hook options, full H2 outline with per-section briefs and word targets, must-include topics, FAQs, internal/external linking suggestions, and a recommended CTA.
Inputs
- Target keyword (required)
- Page type — Blog / Review / Guide / Listicle / Product
Outputs
- Differentiation angle
- 3 title options + 1 meta description
- 3 hook options for the opening
- Full H2 outline with per-section briefs + word targets
- Must-cover topics list + must-include keywords
- FAQ questions to answer
- Internal + external linking opportunities
- Recommended CTA
- One-click Markdown export
Pro tips
- Match the page type to the SERP. If SERP X-Ray said
"dominant format: listicle", pick Listicle here. Don't pick Blog and write a flowing essay. - The Markdown export feeds straight into Site Engine, Sales Page Engine, or your own writing tool. No retyping.
- Use the brief as a contract — if you outsource writing, hand the writer the markdown and tell them to hit every H2 + FAQ. Your editing time drops 60%.
09 Recipe workflows
Five real workflows showing how to chain the tools together for specific outcomes.
Recipe 1 — Launch a new niche site
- Run Keyword Research on your broad niche seed (3 cr).
- Note the 6 topical clusters returned. Each one = a piece of content.
- For your top 3 clusters, run SERP X-Ray on the cluster's lead keyword (12 cr total).
- Pick 12 long-tails from the keyword table — that's your first 12 articles.
- Run Content Brief on the first 3 priorities (18 cr).
Total cost: ~33 credits for a full launch plan + 3 ready-to-write briefs.
Recipe 2 — Outrank a specific competitor
- Run Competitor Snap on their ranking page (4 cr).
- Read the weaknesses + outrank strategy sections.
- Run SERP X-Ray on their primary keyword to confirm format + intent (4 cr).
- Run Content Brief with their primary keyword + the page type the X-Ray recommended (6 cr).
- Write the piece, exploiting every weakness Claude flagged.
Total cost: 14 credits to outrank an established page.
Recipe 3 — Diagnose why your page isn't ranking
- Run Page Optimizer with your URL + target keyword (3 cr).
- Work the critical + high fixes first.
- Run SERP X-Ray on the same keyword (4 cr) — confirm you're targeting the right format.
- Run Competitor Snap on whoever's at #1 (4 cr) — copy what they're doing right.
- Republish. Re-run Page Optimizer in 2 weeks (3 cr).
Total cost: 14 credits + 1 follow-up audit.
Recipe 4 — Build a content cluster
- Run Keyword Research on the sub-niche (3 cr) — pick the densest cluster from the output.
- List 10 keywords from that cluster. Each becomes a piece.
- Run Content Brief on the pillar keyword (the broad one) — that's your hub article (6 cr).
- Run Content Brief on 4–6 supporting keywords — those are your spokes (24–36 cr).
- Write all of them. Internally link spokes → hub and hub → spokes.
Total cost: 33–45 credits for a full topical cluster ready to write.
Recipe 5 — Pre-flight before writing a sales page
- Run SERP X-Ray on your money keyword (e.g.
best ai writing tool) (4 cr). - Confirm intent is commercial. If it's informational, you've picked the wrong keyword.
- Run Competitor Snap on the #1 sales/review page (4 cr).
- Run Content Brief with page type = Review or Listicle (6 cr).
- Hand the markdown brief to the Sales Page Engine (or write it yourself).
Total cost: 14 credits before you write a single line of sales copy.
10 Credits & history
Cost summary
| Tool | Cost per run | Refund on failure? |
|---|---|---|
| Keyword Research | 3 cr | Yes — auto-refunded |
| SERP X-Ray | 4 cr | Yes — auto-refunded |
| Competitor Snap | 4 cr | Yes — auto-refunded |
| Page Optimizer | 3 cr | Yes — auto-refunded |
| Content Brief | 6 cr | Yes — auto-refunded |
Credits are deducted up-front, before the run starts. If the run fails (timeout, parse error, blocked URL), credits are auto-refunded within seconds — you'll see your credit pill rebound after the error toast.
Run history
Every successful run is saved to your Run History tab. The last 100 runs across all 5 tools are kept. Click any history row to instantly reload the full result — no credits charged on replay.
11 Limits & edge cases
What works well
- English-language keywords (US, UK, India, Canada, Australia, Global)
- Public HTML pages — blogs, sales pages, review sites, ecommerce product pages
- Short-tail and long-tail keywords (
"shoes"and"best running shoes for flat feet 2025")
What's flaky or unsupported
- JS-rendered SPAs — some Webflow / Notion / Next.js client-rendered pages return thin content. Pick HTML-rendered competitors when possible.
- Paywalled / login-required URLs — the scraper sees the wall, not the article.
- Non-English SERPs — Claude can analyze them but quality drops. Best results in English.
- Sites that block bots — Cloudflare-heavy sites sometimes 403 the scrape; you'll see a clear error and credits will be refunded.
- Local-pack / map-pack queries — we read organic results only, not the map pack.
Timing
Each tool has a 12-second fetch timeout. If a target site is slow or unresponsive, you'll get a clean error — no hanging requests. Total run time is fetch time + Claude processing (~5–15 seconds for the AI itself).
12 Frequently asked questions
Different scope. Those tools sell you databases — billions of pre-crawled keywords with cached metrics. They cost $99–199/month because the database is the product.
This engine sells you live decisions. We fetch fresh SERPs and pages on every run, then Claude reads them and tells you what to do. No stale database, no monthly subscription floor — you only pay credits when you actually use it.
Pick this if you want a focused 5-tool decision workbench. Pick Ahrefs if you need to export 50,000 keywords at once for an enterprise content team.
Because nobody knows the exact volume. Google stopped publishing real-time keyword volume to non-advertisers in 2014. Every SEO tool that shows you "1,247 searches/month" is using statistical estimation models that have wide error bars.
We use buckets because they're honest and decision-useful. low + commercial is a green-light keyword regardless of whether it's "300/mo" or "900/mo".
Google blocks scraping aggressively and the official Google Search API is expensive + rate-limited. DuckDuckGo's HTML index pulls from the same web and surfaces ~95% the same top results for English queries — at no cost, no key, no rate limit.
If you've been spending $50/mo on a SERP API to get the same data, this is the structural reason this tool can include unlimited SERPs in your Premium plan.
Yes — every tool refunds credits automatically on failure. You'll see a clear error toast and your credit pill will rebound to the pre-run balance within a couple of seconds.
Yes. The output is yours — research keywords for your clients, brief their writers, audit their pages. The runs are scoped to your account so client A and client B's work stays separate in your history.
If you want a resellable white-label PDF deliverable, that's the Elite SEO Audit Engine (separate engine in your dashboard).
Typical workflow for one publish-quality article:
- 1× SERP X-Ray (4 cr)
- 1× Competitor Snap on the #1 (4 cr)
- 1× Content Brief (6 cr)
Total: 14 credits per article. Add 3 cr if you also want to run Page Optimizer after publish.
Yes:
- Keyword Research → Copy as CSV (clipboard, paste into Sheets / Excel / Notion).
- Content Brief → Copy as Markdown (paste into your CMS, Site Engine, Sales Page Engine, ChatGPT, anywhere).
- SERP X-Ray / Competitor Snap / Page Optimizer → screenshots or copy-paste from the rendered output (a structured-data export is on the roadmap).
The page is most likely client-rendered with JavaScript. Our scraper reads HTML as the server delivers it — it doesn't execute JS. Sites built on Next.js, Webflow, Notion, or React SPAs that render content client-side will return only their pre-render skeleton.
Workaround: pick a different competitor that uses server-rendered HTML, or treat the snapshot as directional rather than precise.
Technically yes — Google Suggest returns suggestions in the seed language and Claude can analyze them. But results are noticeably weaker for non-English. We're focused on English (US/UK/India/Canada/Australia) for now.
No — they're complementary. Think of it this way:
- Elite SEO Engine (this one) = your daily research workbench. You use it 3–5x a week to plan and write content.
- Elite SEO Audit Engine = your client deliverable. You use it 1–4x a month to crawl a client's site and produce a 30-page white-label PDF you bill for.
Stuck on something?
If a run keeps failing or a result looks off, open a ticket — we usually patch quickly.