Keyword Research

Most keyword research is manual, slow, and shallow. Agencies charge monthly retainers to run the same Semrush export you could run yourself. We built a pipeline that uses AI to discover, cluster, and prioritise keywords at a depth and speed that manual processes cannot match. For a $9B client, it reduced keyword research from 3 months to 15 minutes.

Get in touchHow it works
How it works

Our process

01

Competitive keyword discovery

We scrape your competitors' ranking keywords, identify gaps between their coverage and yours, and surface the terms where you have the best chance of ranking. Not a Semrush CSV dump. A structured analysis of where your competitors are weak and you're not showing up.

02

AI-powered clustering

Raw keyword lists are useless without structure. Our pipeline groups keywords into topical clusters using embedding similarity, not just string matching. This means semantically related terms get grouped together even when they share no words. The output is a cluster map that directly informs your site architecture and content strategy.

03

Search intent classification

Every keyword gets classified by intent: informational, navigational, commercial, transactional. This determines what type of page you need for each cluster. A blog post won't rank for a transactional query. A product page won't rank for an informational one. Intent mismatch is the most common reason good content doesn't rank.

04

Content briefs

Each cluster produces a brief: target keyword, supporting keywords, intent, recommended page type, competitor pages currently ranking, and content gaps to exploit. These are ready to hand to a writer or feed into a content pipeline.

Deliverables

What you get

Full keyword universe for your domain and competitors
Clustered keyword map organised by topic and intent
Gap analysis showing where competitors rank and you don't
Content briefs for priority clusters
Recommended site architecture changes based on keyword data
Results

Proof, not promises.

$9B enterprise client

Keyword research process reduced from 3 months of analyst time to 15 minutes of pipeline runtime. The output was more comprehensive than the manual process because the pipeline discovers keywords that manual research misses.

Pricing
Keyword research engagements start from $3,000 for a single-domain analysis with clustering and content briefs. Multi-domain competitive analysis and ongoing keyword monitoring are scoped individually. Get in touch to discuss your requirements.
Related services

Competitor Analysis

Pair keyword research with a full competitive audit

Content Strategy

Turn your keyword clusters into a publishing roadmap

SEO Consulting

Implement keyword findings across your site

Embeddings & Similarity Search

The ML behind our clustering pipeline

FAQ

Frequently asked questions

How long does keyword research take?

Our AI pipeline completes discovery, clustering, and brief generation in hours, not weeks. A typical single-domain analysis with competitor gap research and content briefs is delivered within 5 business days, including review and refinement.

How is AI keyword research different from using Semrush or Ahrefs?

Tools like Semrush and Ahrefs provide raw keyword data. Our pipeline goes further: it clusters keywords by semantic similarity using embeddings, classifies search intent, identifies content gaps against competitors, and produces structured briefs ready for content production. The tool is the starting point, not the deliverable.

What industries does your keyword research cover?

We work across industries. Our pipeline is domain-agnostic — it analyses whatever competitive landscape your business operates in. We have experience in ecommerce, SaaS, professional services, and healthcare.

Do you provide ongoing keyword monitoring?

Yes. Keyword landscapes shift as competitors publish and Google updates its algorithms. We offer monthly monitoring retainers that track ranking changes, surface new keyword opportunities, and update your cluster map as the market moves.

What is keyword clustering?

Keyword clustering groups related search terms into topics based on semantic similarity, not just shared words. This tells you which keywords can be targeted by a single page and which need their own dedicated pages. Proper clustering prevents keyword cannibalisation and informs site architecture.

Get started

Send us your domain and your top three competitors. We'll come back with an initial keyword gap snapshot so you can see what the pipeline finds before committing to a full engagement.

Get in touch →