Jump to

Skill · SEO rank tracking

SEO rank tracking.

Segmented keyword sets produce signal; random lists produce noise.

Set up and run an ongoing rank tracking program: pick the right keywords, segment them by purpose, baseline the starting state, and set alert thresholds that filter noise. A good tracked set has structure, grouped into four buckets, because a random keyword list produces a noisy dashboard while a segmented one produces signal.

It runs on Ahrefs MCP and Rank Tracker data. Most properties track 200 to 500 keywords; the discipline is quality over volume and a baseline captured on day one.

Audience: SEO teams starting tracking for a property, baselining before a campaign, or building a ranking dashboard and reporting cadence for stakeholders.

View the skill on GitHubBrowse the full catalog

The framework

Four buckets give the tracked set structure.

Group every tracked keyword into one of four buckets. The structure is what turns a keyword list into a readable dashboard.

  1. 01Brand keywords: queries with the brand name and its variants and misspellings. Typically 10 to 30; the healthy signal is position 1 for every variant.
  2. 02Money keywords: high-intent transactional and evaluation queries tied to revenue. 30 to 100; the healthy signal is the top 3 to 5, watched for drift in either direction.
  3. 03Opportunity keywords: queries ranking on page 2 to 3 (positions 11 to 30) with upside into the top 10. 50 to 200; the healthy signal is trending toward page 1 over 90 days.
  4. 04Competitor benchmark keywords: a representative sample of shared head and torso queries, used to track relative position and share of voice. 50 to 150.

Filter the noise

Default thresholds, one per bucket.

Not every position change deserves attention. Thresholds per bucket filter noise, and loose ones cause alert fatigue until stakeholders stop reading.

  1. 01

    Brand

    Alert if any tracked brand variant drops below position 1. Routed to the SEO lead and brand manager immediately.

  2. 02

    Money

    Alert if a position drops by 5 or more places, or falls out of the top 10. Routed to the SEO lead in a daily digest.

  3. 03

    Opportunity

    A positive alert when a position rises into the top 10, and a drop alert past position 30. Routed to a weekly summary.

  4. 04

    Competitor benchmark

    Alert when aggregate share of voice changes by 10 percent or more. Routed to a monthly summary.

Runs on this platform

The program's data backend.

The program runs on the Ahrefs MCP and Rank Tracker. The existing ranked keywords come from Ahrefs Site Explorer, and the tracked positions, SERP composition, and competitor overlap all come from the Ahrefs integration.

Reference files

The reference that goes alongside the SKILL.md.

  • references/dashboard-template.md

    A template for the rank tracking dashboard layout, including the recurring report structure for weekly and monthly cadences.

Browse all reference files on GitHub

Bridges to other skills

What feeds the tracker, and what reads it.

Tracking is ongoing monitoring. These cover the discovery that fills it, the audit that feeds the opportunity bucket, and the diagnosis when a tracked position moves.

  • Discovery

    seo-keyword

    Finds and classifies the queries worth tracking. Research builds the universe; this skill picks the subset worth monitoring closely.

  • Fills the opportunity bucket

    seo-keyword-gap-audit

    The gap audit's quick-wins list feeds the opportunity bucket directly. Tracking then watches whether those bets move toward page 1.

  • When a position moves

    seo-traffic-diagnosis

    When a tracked ranking has already dropped and the question is why, that is diagnosis. Tracking raises the alert; diagnosis finds the cause.

  • Single-page work

    seo-onpage

    Optimizing one page to move it up the tracked set is on-page work. Tracking shows which keyword earns the effort.

Open source under MIT

Read the SKILL.md on GitHub.

The skill source lives in the rampstackco/claude-skills repository alongside dozens of other skills covering the full lifecycle of brand and product work. This page is a structured overview; the SKILL.md is the source. MIT licensed.

View SKILL.mdBrowse the full catalog

Frequently asked questions.

What should I track, and how much?
Most properties track 200 to 500 keywords, segmented into the four buckets. Tracking too many (a 5,000-keyword set) produces an unreadable dashboard, and tracking too few (30 keywords) misses the picture. The principle is quality over volume: 50 well-chosen money keywords beat 500 noisy ones. Drop dead keywords and add new ones over time, and rebuild the opportunity bucket quarterly because opportunity is a moving target.
What are the four tracking buckets?
Brand keywords (queries with the brand name and variants, typically 10 to 30, where anything below position 1 is investigate-worthy), money keywords (high-intent revenue queries, 30 to 100, watched for the top 3 to 5), opportunity keywords (page-2-and-3 rankings with upside, 50 to 200, watched for the trend toward page 1), and competitor benchmark keywords (shared head and torso queries, 50 to 150, used for share of voice). Each bucket has a different healthy signal and a different review cadence.
How do I avoid alert fatigue?
Set thresholds by bucket so only meaningful changes alert: a brand variant dropping below position 1, a money keyword dropping 5 or more places or leaving the top 10, an opportunity keyword breaking into the top 10 or falling past 30, and competitor share of voice shifting 10 percent or more. Route by urgency (brand and money fast, opportunity weekly, competitor monthly). Loose thresholds produce too many alerts until stakeholders stop reading them, so tighten thresholds in high-volatility niches.
Why capture a baseline?
Because without day-one positions and SERP composition, every dashboard is just current state, and the question 'did the campaign work' becomes unanswerable. Record the current position, the SERP composition (top 10 domains, dominant content type, SERP features present), the current click-through rate at that position, and the date for each keyword. The baseline is the reference everything else is measured against, so capturing it before any change is non-negotiable.
Is position enough to report?
No. Conflating position with traffic is a failure mode: a position-1 keyword with no clicks is worse than a position-5 keyword with steady traffic, so layer in CTR and clicks rather than reporting rank alone. Always show movement by bucket rather than as an aggregate average, which can hide important shifts, and report with context: 'up from position 11 over 60 days' is meaningful, while 'we are position 4' on its own is not.