Homearrow_forwardServicesarrow_forwardSEOarrow_forwardKeyword Research
05SEO
Find real customer queries—and map them to pages
Cluster intents, map them to money pages, cut near-duplicates, and sequence the first 90 days around what can convert.
Category / SEOEstimated timeline / 1–2 weeksBudget / Quoted after we review markets and services
Scroll to continueEN / 2026
Teams that chase volume alone often pick hard head terms or queries that never convert. Useful research starts from sellable services and real customer language.
A strong fit when
This service fits when
Teams starting SEO without a shared keyword list
Content plans that need profitable topic choices
Sites with cannibalization from unmapped queries
Expected outcomes
What the work should make clearer
Money vs support queries are clear—near-duplicate pages drop
A keyword↔page map exists with a first-90-day order
Writers update pages by intent—not search volume alone
Quick summary
What you get when you hire this service
05
01Keyword clusters by intent (learn / compare / buy)
02A keyword ↔ target-page map
03Cycle-one money pages plus URLs to merge or drop
04Rules for not creating a new page when intents collide
05Metrics per page group after shipping
From clients
What people who worked with us notice
“The intent map locked course/schedule URLs as money pages—trials +56% and time to find schedules ~−50%.”
“Mapping queries to revenue categories first stopped us rewriting every SKU in one cycle.”
Related work
Example work clients can review
Swipe through cases close to your brief. Each one focuses on a measurable outcome—not looks alone.
When keyword research should come before content production
It pays off before a large set of service/article pages, or when near-duplicate keyword pages already collide.
If a campaign list already exists, this work is often intent cleanup and page mapping—not a full rediscovery.
Fictional education scenario: course and article pages overlap. Use an intent map to lock course and schedule URLs as money pages, then measure trial requests from real data.
Fictional e-commerce scenario: map queries to revenue categories first so store SEO focuses category and product templates instead of rewriting every SKU in one cycle.
No shared keyword map yet
Lots of content with unclear target pages
You need near-term winnable terms—not only head terms
02
Business inputs to lock before research
Useful keywords are filtered by what you sell, where you serve, and what the team can deliver.
Common symptoms: high-volume lists with no target URL, or several pages competing for the same term. Strong work ends with clusters mapped to URLs and a first-round order.
01
Inputs that make the list actionable
Without these, lists get too wide to prioritize.
Core services/products and offers
Geography or audience constraints
Phrases customers use in chat or calls
Important existing URLs
Combine real customer language with search data before clustering and page mapping.
03
What keyword research delivers
Delivery should be a file content and SEO can action immediately.
A good handoff is not word count—it is knowing which terms are realistic first and which URL owns each cluster, without rank promises.
Clusters with intent notes
Comparative difficulty/opportunity notes
Mapping to current or proposed URLs
A 4–8 week first-cycle order
A useful map states both the query and the page that owns it.04
Research order and anti-patterns
After the draft list, business review must cut queries you cannot serve—wrong region or off-offer work.
Skipping a business review often produces content from terms you cannot serve, or accidental duplicate pages.
01
Steps
From collection to page mapping.
Collect terms from business, competitors, and tools
Cluster by topic and intent
Score opportunity against resources
Map URLs and sequence cycle one
02
Checks before handing to content
Avoid articles with no owning page.
Every core cluster has an owning URL
No accidental multi-page fights for one query
Chosen terms match sellable services
Sequence does not always start with the hardest head term
