/keyword-deep-dive

Difficulty From SERP Signals

How to classify a keyword as Easy, Moderate, or Hard by reading the actual SERP instead of trusting a KD score from a paid tool.

Why We Reject KD Scores

KD scores from Ahrefs, Moz, and Semrush are modeled estimates built from different formulas. Ahrefs leans primarily on referring domains; Semrush blends backlink strength with keyword characteristics. There's no industry-standard formula, and Keywords Everywhere's comparison notes the tools can differ by 30+ points on the same keyword. A KD of 15 in one tool can be a KD of 55 in another. That's not a minor disagreement; it's a signal that the metric isn't measuring what you care about.

What you actually care about: can a site like mine realistically earn a page-1 position in the next 6-12 months? That's a question about the specific SERP and your specific position on it, not a single number. The rubric below replaces the KD score with a signal count you can read from the live SERP in 10 minutes.

Easy Signals

• Top 10 dominated by small sites (no recognizable publisher logos)
• Thin content in the top 5 (pages under 1,000 words with minimal structure)
• Only 3-4 pages actually answer the query; positions 5-10 are loose matches
• Clean blue-link SERP with no featured snippet or AI Overview
• No Wikipedia or government domain in top 10
• Top 3 have no visible E-E-A-T signals

Most of these present? A modest-authority site can plausibly reach top 5 in 3-6 months with a single well-executed page.

Moderate Signals

• Mix of authority and small sites in top 10 (HubSpot alongside a small blog or a forum thread)
• 1-2 SERP features present but not dominating
• Content depth varies (800-word pages alongside 3,000-word pages), meaning Google hasn't locked in a depth preference
• Top 10 Domain Rating roughly mixed in the 30-60 range

This is where the game is won with a better page than the mid-pack incumbents. Expect 6-12 months to top 5 with sustained execution and internal linking from your existing cluster.

Hard Signals

• Every top 10 is a high-authority site (major publishers or well-known category leaders)
• Top 10 content consistently 2,500+ words with strong structure
• Multiple SERP features stealing clicks (featured snippet, AI Overview, image pack stacked on one SERP)
• Original data or proprietary frameworks in top 3 that you'd have to build from scratch
• Every top 5 has named authors with credentials plus first-party research artifacts
• A dedicated niche-authority site holds a top 3 position

Expect 12-24 months before top 5 is realistic. A one-page play won't work; you'll need a cluster plus authority growth, usually backed by original research.

The 10-Point Rubric

Count each signal that applies. Sum the points and map to a bucket.

1. Top 10 includes Wikipedia or a government/institutional domain. (+1)
2. Top 10 has more than 5 high-authority publishers (HubSpot, Forbes, NYT, major SaaS brands). (+1)
3. Every top 5 result is over 2,500 words. (+1)
4. Featured snippet is held by a top-3 authority brand. (+1)
5. AI Overview is present on the SERP. (+1)
6. Every top 5 has named authors with visible credentials. (+1)
7. Top 3 include original data or proprietary frameworks. (+1)
8. Top 10 has consistent freshness signals indicating active maintenance. (+1)
9. A dedicated niche-authority site holds a top 3 position. (+1)
10. Meta descriptions in top 5 are clearly CTR-optimized. (+1)

Mapping:

• 0-3 = Easy. A focused new page can plausibly reach top 5 in 3-6 months. Single-page play is viable.
• 4-6 = Moderate. Cluster approach recommended. 6-12 months to top 5. Requires better content than mid-pack incumbents.
• 7-10 = Hard. 12-24 months. Requires original research plus sustained authority building.

The rubric is directional, not mathematical. If you're between two buckets, pick the harder one and let reality correct you.

"Shouldn't Attempt" Signals

Some SERPs are unwinnable regardless of rubric score. If any of these apply, the honest answer is "skip this keyword."

• Wikipedia holds top 3. You're not outranking Wikipedia on general-knowledge queries. Target a long-tail variation.
• Reddit dominates top 10. Reddit threads rank because users explicitly want discussion format. Google's "hidden gems" update cemented Reddit's position on many queries. You can't replicate a forum discussion with an article.
• Government or official organization dominates. On regulations or health guidelines, Google trusts the source of authority rather than content quality. Outranking is not realistic.
• Pure news cycle SERP. See serp-volatility-heuristics.md Level 4. These belong to Google News publishers.
• AI Overview displaces most clicks AND the query is informational AND you can't get cited inside the AIO. Click potential is too thin. See zero-click-and-aio-impact.md.
• Top 3 is all branded product pages on a non-brand query. Google has decided the keyword is transactional and your editorial content won't rank. Either build a product or skip.

How to Apply This

After reading the top 10, walk through the 10-point rubric and score. Check the "shouldn't attempt" signals; if any apply, flag and recommend pivoting. Map the score to Easy / Moderate / Hard and use it for the timeline estimate in Step 6 of the main skill. If the SERP signals something the rubric misses, override it and explain why. The rubric is a shortcut, not a rule.