Not how hard. Not how many. Who.
Find the keywords your buyer types.
Ahrefs and Semrush rank keywords by volume and difficulty. ICP Keywords scores every keyword against the person you sell to, your Ideal Customer Profile, then finds new ones in real Reddit threads, YouTube comments and AI assistant answers that no crawled index contains.
7-day trial for $1 — then $29.90/mo, cancel anytime.
- 7 discovery methods
- Opportunity scoring
- SERP clustering
- Video gap
- Sitemap coverage
- Pay-per-use credits
Everyone has the same list
Every tool exports the same crawled index. The keywords worth having are in Reddit threads, YouTube comments and buyer conversations, and no database has them yet.
The number says easy. The SERP says Amazon.
Difficulty counts backlinks to the ranking pages. It never asks who is searching, or whether they would buy. ICP relevance and the live top-3 domain rating sit inside the score.
The post ranked. The AI Overview ate it.
Ahrefs' own study puts the click loss at 58% for position one when an AI Overview is present. The score penalises it, only when it sits above the fold, and weights the commercial searches your buyer makes.
Volume-first tools rank the wrong keywords first
Traditional keyword databases sort by how many people search a term. That is useful for media sites and misleading for products. The biggest term in your niche is often the one your buyer never types, and the term that closes deals sits at 40 searches a month.
This one starts from the buyer
You describe your Ideal Customer Profile once. Every keyword, wherever it came from, is then judged against that profile, and relevance sits inside the Opportunity score next to difficulty, live SERP authority and SERP features. Low-volume, high-fit keywords stop hiding at the bottom of the list.
A look inside the workspace
Real screens from the app. No mockups.
Tune the score to your site, not ours
Set how much ICP relevance, intent, CPC, difficulty and live SERP domain authority each count for. Weights save to your account and apply to every keyword table.
- Separate weighting profiles for runs with and without an ICP
- Live top-3 domain ratings pulled per keyword
- Video results and above-the-fold AI Overviews penalise the score
Group keywords by shared search results
Clustering compares the real top-10 URLs behind each keyword, so terms that Google already treats as one page land in one cluster, with an AI-written page angle and intent label.
- Clusters built from shared ranking URLs, not string similarity
- One row per cluster on export, members pipe-joined
- Up to 500 SERPs per run
See which keywords YouTube hasn't answered
Paste keywords and each one is checked against the live YouTube top 10, then judged as a complete gap, partial gap or covered, with the reason and the closest competing video.
- Complete / partial / no gap triage with rationale
- Flags outdated or low-view incumbents worth outranking
- Filter and export the gaps only
Every run saved, reloadable, exportable
Generated and analyzed runs are stored automatically with the ICP used and the seed they came from. Reload any run with its paid metrics intact. You are never charged twice for the same data.
- Generated, analyzed, clustered and video-gap histories
- Load with metrics, send to paste, or export CSV
- Long opportunity runs can finish in the background
Seven ways to find keywords a database can't
Each method produces keywords from a different source, so you are not re-shuffling the same crawled index everyone else exports.
ICP Discovery
Turns your saved Ideal Customer Profile into the searches that buyer actually types. Jobs-to-be-done and pain points, not synonyms of your product name.
Proxy Intent
Finds the adjacent problems your buyer searches before they know your category exists, so you can meet them earlier in the journey.
pSEO
Builds scalable, data-rooted page patterns and question keywords you can template across hundreds of pages.
Productority YouTube
Reads real comments under videos in your niche and hands back the phrases buyers use, split by how close they are to buying. A one-line product description tells it what to ignore.
Productority Reddit
Same treatment for real Reddit threads. Natural-speech questions people are already asking, never invented keywords.
Role-Play
Puts an AI in your buyer's seat and captures how they would describe the problem, in their words, at each stage from unaware to ready to buy.
LLM Keywords
Shows the follow-up queries ChatGPT, Gemini and Perplexity actually run when someone asks about your topic, with a consensus filter across models.
Then decide what is worth writing
Discovery is the easy half. These tools turn a long list into a shortlist of pages.
Opportunity score
Combines ICP relevance, keyword difficulty, the domain authority of the live top 3, intent and SERP features. Video results and above-the-fold AI Overviews penalise the score, with weights you control in Settings.
Keyword clustering
Clusters keywords by shared URLs in the live top 10, so you know exactly how many terms one page can win. Exports as one row per cluster.
Video gap
Checks each keyword against the top YouTube results and labels it no gap, partial gap or complete gap. A video content plan straight out of your keyword list.
Coverage gap
Indexes your sitemap and matches every keyword against what you have already published, so you stop rewriting pages you own and see the real gaps.
Background cloud runs
Send a large scoring pass to the cloud and close the tab. Progress shows in History and the finished run lands in Analyzed runs.
Run history
Generated, Analyzed, Clustered and Video gap runs are all saved with the ICP used, and can be reloaded with their saved metrics without being charged again.
How a research session runs
- 1
Discover
Pick a discovery method and generate keywords from your ICP, real conversations or AI assistants.
- 2
Size
Pull volume, CPC, difficulty and intent for up to 2,000 keywords at a time.
- 3
Score
Run the Opportunity check for live SERP authority, SERP features and ICP relevance.
- 4
Decide
Cluster into pages, check video and sitemap coverage, then export the shortlist.
ICP Keywords vs Ahrefs vs Semrush
This compares research workflow capabilities, not index size. Ahrefs and Semrush remain excellent all-round SEO suites. If you need a backlink index or rank tracking, keep them. The difference is what happens between "here are 900 keywords" and "here is the page I should write".
| Capability | ICP Keywords | Ahrefs | Semrush |
|---|---|---|---|
| Keywords scored on fit with your Ideal Customer Profile | Every keyword carries an ICP relevance score | Not offered | Not offered |
| Where the keyword ideas come from | Real Reddit threads, YouTube comments, AI-assistant answers and your own ICP | Crawled keyword index | Crawled keyword index |
| Fan-out queries from ChatGPT, Gemini and Perplexity as a keyword source | Built in as the LLM Keywords method | Brand monitoring only, not a keyword source | Brand monitoring only, not a keyword source |
| Keyword extraction from Reddit threads and YouTube comments | Productority Reddit and Productority YouTube | Not offered | Not offered |
| YouTube video gap analysis on your keyword list | Paste a whole list, per-keyword verdict with the incumbent video | Not offered | App Center add-on, one topic per search |
| Coverage gap against your own sitemap | Sitemap indexed and matched against every keyword | Not offered | Not offered |
| Shared-URL keyword clustering | Clusters from live SERP overlap, one page per cluster | Yes | Yes |
| SERP penalties for video and above-the-fold AI Overviews | Applied inside the Opportunity score, weights you control | Not offered | Not offered |
| Backlink index and rank tracking | Not offered | Yes | Yes |
| Billing model | Pay-per-use credits, no seats | Subscription with seats | Subscription with seats |
Testimonials
What users are saying
“I'd never done any SEO for my SaaS before. I took keywords straight from ICP Keywords and published 5 articles. I didn't build a single backlink. Within about 8 weeks they had 2,730 impressions and 65 clicks from Google, and my best article alone brought in 30 clicks. It was the first organic traffic I'd ever had.”
“I used ICP Keywords for my client, Sydney Microblading, and discovered 20+ untapped long-tail keywords around cosmetic eyebrow tattooing, including “eyebrow tattoo for sparse eyebrows,” that I would never have considered targeting. These insights helped me create highly relevant, search-intent-focused articles that are already driving valuable organic traffic to the business. So far, I’ve published 5 articles, and 3 of them reached the first page of Google in less than 6 days! A genuinely useful tool for uncovering SEO opportunities!”
“Using ICP Keywords' LLM Keywords and Productority YouTube/Reddit tools, I generated 137 keyword entries across three runs for research in finance and retirement: 70 from LLM Keywords, 37 from YouTube, and 30 from Reddit. Including three Paste List runs, my history now contains six saved runs totaling 160 keyword entries. I appreciate being able to explore different discovery methods and revisit the saved results while deciding which content ideas to validate further.”
“ICP Keywords has been a useful tool for expanding my keyword research and discovering new content opportunities. Using Find Keywords, I generated 215 keyword candidates from two seed searches, with another 20 entries added through Paste List, giving me 235 keyword entries across three saved runs. I particularly appreciate the ability to save research history, review keyword metrics, and return to previous runs without starting from scratch. These features make it easier to organize research and identify which opportunities deserve further investigation.”
“ICP Keywords has helped me take a more systematic approach to SEO research in the competitive U.S. finance and retirement space. Using Find Keywords and Opportunity Score, I generated 1,083 keyword candidates in one discovery run and another 195 in a separate project. The platform has helped me uncover overlooked search opportunities, organize content ideas, and build a more structured keyword evaluation process. It's become a valuable part of my research workflow.”
Priced by what you run
Research is spiky: some weeks you plan a whole content calendar, some weeks you touch nothing. Credits are consumed per action, so an idle month costs you an idle month. Keyword generation is free and unlimited; credits are only spent on paid data lookups.
$1 trial
$1
100 credits for 7 days for $1 — then $29.90/mo, cancel anytime
- Every discovery method included
- Opportunity Score, Clustering, Video Gap, Coverage Gap
- Run history saved
Standard
$29.90/mo
billed monthly · 3,000 credits/month
- Every discovery method, unlimited keyword generation
- Opportunity Score, Coverage Gap, Clustering, Video Gap
- 5 saved ICP profiles
- 3 indexed sitemaps
- 90 days run history
- CSV exports
Premium
$79.90/mo
billed monthly · 10,000 credits/month
- Everything in Standard
- Unlimited saved ICP profiles
- 15 indexed sitemaps
- Unlimited run history
- Priority support
Agency
$199/mo
billed monthly · 30,000 credits/month
- Everything in Premium
- 30,000 credits per month
- 50 indexed sitemaps
- Built for agencies and heavy users
What the credits buy: 3,000 credits cover about 3,000 opportunity-scored keywords, or roughly 545 keywords through the full pipeline. The Agency plan's 30,000 credits cover about 5,450 full-pipeline keywords. Credits reset monthly and do not roll over.
Pay as you go
Top up any time. Credit packs stay valid for 12 months, and subscribers get 10% off.
Starter
$25
2,000 credits
Growth
$60
5,000 credits
Scale
$100
10,000 credits
Agency
$250
50,000 credits
Questions
- Is ICP Keywords an Ahrefs alternative?
- For keyword research and prioritisation, yes. You get search volume, CPC, keyword difficulty, live SERP top-3 with domain authority, SERP feature detection and clustering, plus an ICP relevance score Ahrefs does not have. It is not a replacement for Ahrefs' backlink index or rank tracker.
- Is it a Semrush alternative?
- For the keyword and content-planning side of Semrush, yes: discovery, metrics, clustering, gap analysis and export. It deliberately skips the advertising, social and site-audit suites so the workflow stays focused on finding keywords your buyers actually search.
- Where does the keyword data come from?
- Search metrics and live SERPs come from a commercial search-data provider. Keyword ideas additionally come from real Reddit threads, real YouTube comments, AI-assistant answers from ChatGPT, Gemini and Perplexity, and your own ICP profile, not only from a crawled index.
- What is an ICP score?
- You save an Ideal Customer Profile: product, category, buyer, pain points and language. Every keyword is then judged against that profile for relevance, and relevance feeds the Opportunity score alongside difficulty, domain authority, intent and SERP features.
- Do I need my own Reddit or YouTube account?
- No. The Reddit and YouTube discovery methods pull public conversations for you; there is nothing to connect or authorise.
- How is billing calculated?
- Credits are consumed per action: keyword metrics, opportunity checks, clustering, video gap and each AI run, so you pay for what you run instead of a per-seat subscription. Runs are saved to history and can be reloaded with their saved metrics without being charged again.
Run your first list against your buyer today
Get 100 credits for 7 days for $1, with every discovery method and every analysis tool unlocked. Card required. Then $29.90/month unless you cancel before the trial ends.