RankingTank
On-Page & Authority · Keyword & Competitor Analysis

Keyword & competitor analysis.

Research that ends in decisions — every term assessed leaves with a verdict and the reasoning that earned it.

TakeBuildLeave
01/ 05 · The research

Competitor website analysis without the wishful thinking

Keyword and competitor analysis is the research that decides what everything else aims at. The data itself stopped being scarce years ago — every tool prints the same columns — so what you're buying here is judgment: which terms are worth taking, which need something built first, and which are traps wearing traffic numbers. We sell the verdicts, with the reasoning attached.

§01 · The researchJudgment

If you want the columns, buy a tool subscription; they're good, and we use one. This service exists because columns don't decide. A 2,900-a-month term with manageable difficulty can still be unwinnable when its results page is wall-to-wall software listicles no service provider will crack — and a 90-a-month term can be the best money on the list because every searcher arrives holding a budget. That call comes from reading the results page, not the spreadsheet. Reading is the work.

02/ 05 · Take / Build / Leave

Every term leaves with a verdict

The deliverable is a target file where nothing stays ambiguous. Each assessed keyword carries one of three verdicts.

FIG.01 · TAKE / BUILD / LEAVETHREE VERDICTS · ONE PER TERMEVERY ASSESSED TERMTakeWINNABLE NOW → ON-PAGE MAPACTBuildSCOPE WHAT MUST EXISTSCOPELeaveNOT WORTH THE FIGHTDECLINE · ON RECORDEVERY ASSESSED TERM LEAVES WITH A VERDICTTAKE · BUILD · LEAVETAKE / BUILD / LEAVE3 VERDICTSEVERY ASSESSED TERMTakeWINNABLE NOW → ON-PAGE MAPACTBuildSCOPE WHAT MUST EXISTSCOPELeaveNOT WORTH THE FIGHTDECLINE · ON RECORDONE PER TERMON RECORD
01 / 03

Take

Winnable with pages you already have. The file names the owning URL, the intent it must satisfy, and what the current page is missing — handed straight into the on-page assignment map, where terms become page specs.

  • On-Page SEO The target file feeds the assignment map directly.
02 / 03

Build

Worth pursuing, but nothing you have can satisfy the intent — the results page rewards a format or asset you don't own yet. The file states exactly what would need to exist, so "build" becomes a scoped decision instead of a standing ambition.

03 / 03

Leave

Not worth the fight: tool-dominated results, misread intent, difficulty priced above the traffic's value, or searchers who will never be your buyers. Leave verdicts save more money than take verdicts make, and we defend them just as hard — on the record, so the debate never reruns in eight months.

03/ 05 · The reading

How the reading gets done

Volume and difficulty are the screen, never the decision. Past the screen, every surviving term gets its results page read by hand: what format wins, who's actually ranking, whether a service provider appears at all, and whether an AI Overview sits on the query — because a term your buyers now resolve inside a generated answer is a different opportunity than one they click through. Competitor gaps get the same skepticism. A competitor's keyword profile is a map of their choices, not your opportunities — half of any gap report is terms they shouldn't rank for either, and we tell you which half. And every take verdict runs a cannibalization pre-check against your existing pages, because giving a term two owners is how rankings rotate instead of climb.

  • GEO We read AI-Overview presence into every verdict.
§03 · The readingBy hand
04/ 05 · What you get

What you get

Deliverables · the target file and its supports
What you get04
  • The target fileevery assessed term with its verdict, intent read, results-page composition note, and owning or proposed URL.
  • Competitor gap analysisagainst the competitors you name and the ones the data surfaces, with the shouldn't-rank noise already removed.
  • Priority sequencetake verdicts ordered by value against effort, build verdicts scoped, leave verdicts recorded with their reasoning.
  • Hand-off without translationthe file feeds the on-page assignment map directly, and flags any term whose ceiling is technical before it's editorial.

One caution we give every client before this engagement: research aimed at a site that can't be crawled cleanly is aim without a rifle. If the audit hasn't happened yet, the targets inherit its findings — sequence accordingly.

  • SEO Audit Targets inherit its findings; sequence the audit first.

Request an audit — or send your current target list, and we'll tell you which verdicts it's missing.

05/ 05 · FAQ

Keyword & competitor analysis FAQ

Isn't this just a Semrush report?

No — the report is the input, and inputs don't carry verdicts. Tools print identical columns for everyone, which is exactly why the columns can't be the edge. The edge is reading each results page for intent, format, and whether a provider can realistically appear there, then binding every term to one owning page with the reasoning written down. A subscription gives you the data. This gives you the decisions, defensible enough to hold up when someone wants to relitigate them.

Do I need this if I'm already getting the audit?

They answer different questions, and most engagements sequence them rather than choose. The audit diagnoses supply — whether your site can be crawled, rendered, indexed, and served fast. This work chooses demand — which queries deserve your pages at all. An audit without targets fixes a site aimed at nothing in particular; targets without an audit aim a site that may not fire. Forced to pick, start with the audit: it finds the ceiling everything else lives under.

Let's build somethingworth finding.

One brief. One reply within a working day. Evidence, not a pitch.

Prefer email? hello@rankingtank.com