How to use KeywordIntent, thirteen workflows for real sites

Quick start, whoever you are

Four steps. Everything after this is a variation on them.

Type a keyword and search. Use the thing you sell or write about, not a clever long-tail phrase. Broad seeds return more.
Look at the keywords that share no words with what you typed. They're badged JUMP. Search running shoes and you get asics gel 1130. No other free tool can reach these, because there's no string to vary, they come from Google's own understanding of what searches mean.
Narrow with the filters on the left until you're looking at a few dozen keywords rather than a few thousand. This is the actual work.
Star the ones you'll genuinely publish and open the Sprint list tab. Ten to twenty is a quarter's work. Export it, or send it to Claude for a running order.

Two buttons cost money to run and are marked as such, Add SEO difficulty and Check page 1. Both show you exactly what they'll spend before they spend it. Everything else is unmetered.

Pick your route

Ecommerce

Your problem isn't finding keywords, it's finding the ones with a buyer behind them, and knowing which product pages are worth building before you build them.

Search your category, not a product. running shoes, not hoka clifton 9 mens size 10.
Turn on “Prioritise commercial intent” in the Scoring block. Opportunity immediately favours buying and comparison keywords over people reading for free.
Filter Type to “Product” and “Buying intent”. The intent jumps here are the prize: brand and model names Google ties to your category that you'd never have typed. Some will be products you don't stock yet.
Sort by Value, not Volume. Value is volume × CPC, what that traffic would cost you in ads. It's the closest thing to a revenue estimate the data supports.
Tick your top ten and hit “Check page 1”. For ecom this matters more than anywhere: the Eff. vol column shows what's left after the shopping carousel takes its cut. A 40,000-volume term with a shopping block above it is not a 40,000-volume term.
Open the Content map tab. Clusters that suggest “Buying guide” or “Review / roundup” are your money pages. Clusters of questions become support content that links to them.

You end with: a category page plan, a shortlist of products worth their own page, and buying guides pointed at both.

Local business

Nobody searches your business name. They search what they want plus where they are, and those phrases barely show up in a normal keyword expansion, because Google returns ideas around a topic, not around a town. You have to build them.

Set the country selector first. It decides which market the numbers come from. A UK town with the country left on United States will show nothing.
Click “Build ‘best … in {city}’ queries →” under the search box.
Enter what you sell (plumber, dog groomer) and your city. Cover several towns by comma-separating them, Austin, Dallas, Houston, and you get a full set for each.
Hit “Analyze these”. One call sizes the whole batch and drops them into your results with a LOCAL badge.
Check page 1 on the ones with real volume. Expect a local pack, the tool shows it in Eff. vol and names it in the expanded row. Where the map pack dominates, your Google Business Profile matters more than a new page.
Upload your Search Console export (“Add my rankings”). Service pages sitting at positions 2–15 get badged LOW-HANGING FRUIT and jump to the top. Improving those beats writing anything new.

You end with: one page per service per town worth having one, ranked by which will actually move.

Google Business Profile

The GBP tools people push at you mostly guess at what to put in your categories, services and posts. Keyword data turns that guess into evidence, which searches actually happen in your town, and which words to put where Google reads them.

Set the country selector to your market first, the numbers come from it.
Click “Build ‘best … in {city}’ queries →” and enter what you do (emergency plumber, invisalign dentist) and your city, comma-separating every town you serve.
Hit “Analyze these” to size the whole batch, they drop in with a LOCAL badge.
Read the words that repeat across the ones with volume. The service names, conditions and “near me” phrasings that keep appearing are your GBP categories, services and business-description keywords, taken from real demand rather than invented.
Check page 1 on the ones worth it. Where the map pack sits above the results, the tool flags it in the expanded row, that is a GBP and reviews job, not a new page. Where it doesn't, the local landing page is worth building.
Star the page-worthy terms into a Sprint list, and turn the map-pack ones into GBP post topics and service entries.

You end with: GBP categories, services and posts chosen from real searches, plus a local page list for the terms the map pack doesn't already own.

Local lead gen (rank-and-rent)

You're not ranking a business, you're building pages that print leads you sell. The whole game is picking service-plus-town combinations with real demand and a page one you can take without any authority behind you.

Set your market, then “Build ‘best … in {city}’ queries →” for the service and every town you'd rent to. Comma-separate a whole county in one go.
Analyze the batch and sort by Value, volume × CPC. That CPC is the closest read you have on what a single lead is worth to the business you'll sell it to.
Check page 1 on the top towns and read Open?. A page one of directories and thin local pages is an open door, a town you can rank a cheap page in this month.
Turn on “Winnable page 1 only” to strip the list down to exactly those towns.
Score against my buyer (Pro) with a buyer profile of the business that buys your leads, so the Fit column ranks towns by the leads that buyer actually pays for, not just by volume.
Star the winnable, high-value combinations, each one is a single page to build and rent.

You end with: a build list of service-town pages ranked by lead value and how cheaply you can rank them.

Niche / authority site

You're building topical depth, so you need the shape of a subject, not a list. And you need to know which corners of it you can afford to compete in.

Search your niche broadly and let Google Suggest finish loading (the row count says “finding more phrasings…” while it works). It adds the questions Ads data has none of.
Switch on “Intent jumps only”. This is the whole reason to use this tool for a niche site: the terms your topic connects to that contain none of your words. That's where topical gaps hide.
Open the Content map tab. Each cluster is a silo, a hub page plus its supporting posts. The suggested format tells you which.
Before committing, click “Add SEO difficulty”. Ignore KD and read Links to beat, the referring domains the pages currently ranking actually have. Two keywords twelve KD points apart can differ twentyfold in the links required. That number decides what a new site can realistically win.
Copy brief on one cluster and build that silo properly before starting another.
Star that cluster's keywords into the Sprint list so the next three months has an end point.

You end with: one silo you can genuinely rank, not forty you can't.

New Site Playbook

A brand-new site has no authority, so it needs keywords where the pages already ranking have few links and page 1 still has room for someone small. This workflow screens your niche for both and gives each keyword a verdict with the reason. Pro: it uses your monthly credits (or your own DataForSEO key), plus your own AI key for the outlines.

Click “⚡ Run a workflow →” under the search box and pick New Site Playbook. Type your niche as a broad keyword (cold plunge, not a long question). If it's 3 words or fewer, the tool first asks “Which do you mean?”, grouping what people actually search (supplements, workouts, questions…) so you can tick the meanings you want. Each one gets its own run. Got several niches? Put one per line and they run as a batch with one approval.
Check the cost, then run it. The tool pulls your niche's keywords, screens up to 200 of the best for Links to beat in one request, then checks page 1 on the 20 easiest. That's at most 22 credits a niche, and you see the maximum before anything is charged.
Read the New site column. Rankable: forums, free-platform blogs, sites under DR 30 or other weak pages already hold spots in the top 10, so a new page can take one. The Lowest DR column shows the weakest site on page 1 (Domain Rating by Ahrefs). Video-led: page 1 is mostly videos, so a video (with a page to match) is your best route in. Possible: few links to beat, but established pages hold page 1, so it needs your strongest page. Hover a verdict, or open a row, to see why. An AI Overview above the results is flagged as a caution, never counted as a plus.
Flip “New-site picks only” in the left rail to hide everything else, and use “↓ Export CSV” above the table to take the list with you.
Want a second opinion? Tick keywords and hit “✦ Recheck ticked with KGR” in the New site panel. It counts the pages using the phrase in their title (allintitle), and in their URL too if you tick “Also count allinurl”. Google only shows about a page of these, so treat KGR as supporting evidence, not the decision.
Hit “✦ Outline articles” in the toolbar. It clusters your picks (one article per cluster), you add a little about your offer (paste sales copy or a URL, and pick a saved buyer if you have one), and your AI drafts the outline. Use “Write the full article in Claude” for the full draft.
Keep watching the niche with “⏰ Repeat this” in the workflow window. It reruns weekly or monthly, and your dashboard shows what's new, what's gone and what's now rankable. Star the winners into your Sprint list so the first month has a finish line.

You end with: a short list of keywords a brand-new site can genuinely rank, each with the reason, and an outline ready to write for each cluster.

Blog / content site

If you've been publishing a while, your biggest wins are already on your site. Google has decided you belong on page one or two for things you've forgotten you wrote.

Start with Search Console, not the search box. Open Performance → Queries → Export → Download CSV, unzip it, and drop Queries.csv into “Add my rankings”. Nothing leaves your browser, the file is read locally.
Search a topic you already cover. Anything you rank 2–15 for gets badged LOW-HANGING FRUIT, with your position in the My pos column, and is pushed to the top of the list.
Flip on “Low-hanging fruit only”. That's your rewrite list. Moving #9 to #4 is an afternoon; ranking something new is a campaign.
Filter Type to “Question” for genuinely new pieces. Questions are where blogs still win, and where Google Suggest fills gaps the Ads data can't.
Send a cluster to Claude with “Copy brief”. Where you rank 50+ for something specific, the brief tags it → NEW DEDICATED PAGE, that's not a ranking problem, it's Google telling you your broad post isn't what those searchers want.
Star 10–20 and work the Sprint list until it's done, then clear it and pick again.

You end with: a rewrite list that pays this month and a publishing list that pays next quarter.

Affiliate site

Every affiliate keyword worth having is contested. The question isn't what to target, it's which contested pages you can actually get onto.

Search the product category and turn on “Prioritise commercial intent”.
Filter Type to “Comparison”, then use the Contains box for vs, alternative, best and review in turn. These are the pages that convert.
Use wildcards to mine a brand. * mailchimp gives you what people search before the name, alternatives and comparisons. mailchimp * gives pricing, complaints and integrations.
“Check page 1” on your shortlist and read the Open? score. A page one full of Reddit, YouTube and Medium isn't crowded, it's open. Google is telling you it will rank content you don't need domain authority to publish.
Expand any row to see exactly who holds page one and which platforms you could publish on today.
Check Links to beat before writing. If the ranking pages carry thousands of referring domains, that keyword isn't a content problem.

You end with: commercial keywords ranked by whether you can win them this year, not just by volume.

SaaS / software

Your buyers search for competitors, not for you. This is the workflow for getting in front of them.

In the search box at the top, put a competitor's domain instead of a keyword (e.g. clearserp.com). The tool detects it's a domain and returns every keyword they rank for, with their position.
Click “Fetch weak spots only →”, the button just above the results table on the right. Positions 5–20 are the ones they hold loosely: position 1–4 is a fight, position 12 is an opening. Tick the ones worth taking and star them into your Sprint list now (the ☆ on each row), because the next step runs a fresh search.
Now run a second, separate search, this time a keyword. This one replaces the table on screen, but your Sprint list keeps everything you starred, so you lose nothing (and the domain stays in your Recent chips to jump back to). Go back to the search box at the top, clear the domain, and type your category as a plain keyword (for a ClearSERP-style tool that's rank tracker or serp api). Search, then flip on “Intent jumps only”, the first toggle at the top of the left-hand rail. You'll get rival product names, the terms Google ties to what you do without containing any of your words.
Mine each rival with wildcards, in the same search box. * clearserp surfaces “alternative to”, “vs” and “cheaper than”; clearserp * surfaces pricing complaints and missing features. Both are pages you should own.
In the left rail, under “Search intent”, tick “Commercial” and “Transactional” to strip out people who are only reading.
Tick the alternatives and comparison rows (the checkbox on the left of each row) and star them into your Sprint list with the ☆ on each row. Open the Sprint list tab above the table to work through them, that's a quarter of bottom-of-funnel content with a defined finish line. Comparing several rivals or running more than one campaign? The Sprint list is a single account-level list, so save each finished campaign to a named Collection (“Save to collection”), Clear the sprint between them, and reopen the right Collection from your dashboard when you come back to that campaign.

You end with: an alternatives-and-comparison plan aimed at people already shopping for what you sell.

AEO / AI search

More of the searches you want now end in an AI answer, not a click. You can't rank inside an AI Overview, but you can be the source it quotes. That means targeting the questions AI answers and writing the passage it lifts.

Search your topic broadly and let Google Suggest finish loading, the real questions live there, and the Ads data alone doesn't carry them.
Filter Type to “Question” and switch on “Intent jumps only”. Those are the questions Google ties to your topic that you'd never have written down yourself.
Check page 1 on the shortlist. Rows badged AI Overview already trigger one, those are the queries where being cited is the whole prize. Where the badge reads AIO ✦ cite-able, the answer quotes platforms you can publish on today.
Read Eff. vol on those rows. An AI Overview takes a chunk of the clicks, so treat these as citation targets, not traffic forecasts.
Open the Content map tab and collapse the questions into one answer page per theme, each with a short, quotable definition near the top.
Send the cluster to Claude with “Copy brief” and ask for a page written as direct answers under question headings, the shape AI engines quote.

You end with: the questions AI actually answers in your space, and a page plan built to be the source it cites.

Info product / course

You're not chasing traffic, you're finding the people with the exact problem your course solves and meeting them at the moment they go looking for the answer you sell.

Add your course as a buyer profile in the dashboard (Pro), what you teach, who it's for, and the pain it removes.
Search the problem it solves, not the course name. content that stops getting traffic, not the name of your workshop. Broad seeds surface the intent jumps.
Hit “✦ Score against my buyer” (Pro). The Fit column ranks every keyword by how well its searcher matches your student, not by how big it is.
Turn on “High fit only” and open “Best 20 for my buyer” from the Pick the best 20 menu, that is your content shortlist.
Open the Action plan (✦). Its buyer-stage section splits the list into TOFU, MOFU and BOFU, the awareness posts, the comparison pieces, and the pages that actually sell.
Star the BOFU and MOFU terms into a Sprint list, they map straight onto lessons and sales-page sections.

You end with: a content plan ordered by who your student is and how close they are to buying, not by volume.

Coaches & consultants

You sell trust and an outcome, not a product. The keywords that matter are the ones your ideal client types once they've realised they need help but haven't picked who, and those hide behind the informational noise.

Add a buyer profile for your ideal client (Pro), their role, their situation, and the result they want.
Search the outcome you deliver and the problem behind it, broadly, so the intent jumps come through.
Score against my buyer, then sort by Fit. The high-fit terms are the ones your actual client searches, even at low volume. Low volume plus perfect fit beats high volume and none.
Filter Search intent to “Commercial”, people weighing up options, and “Transactional”, people ready to book.
Check page 1 and read Open?. Where forums and directories rank, one clear authority page or a comparison piece can outrank them.
Open the Action plan for the buyer-stage split and star the BOFU terms, those become your services pages and case studies.

You end with: a short, high-fit list aimed at clients who are close to hiring, not a long list of readers.

Agency / client work

Two jobs: win the client, then show them something finishable. This tool does both without a seat licence.

For the pitch, run their domain and their competitor's. “Fetch weak spots only” on the competitor gives you a slide of keywords the prospect is losing that they could take.
Ask for their Search Console export and upload it. The LOW-HANGING FRUIT list is the fastest credibility you will ever build, wins they already half-own, that you can show moving inside a month.
Use the Sprint list as the engagement. Ten to twenty keywords is a quarter of work with a defined end, which is a far easier thing to sell than “ongoing SEO”.
Export the sprint as its own CSV for the proposal, and the full table for the appendix.
Use “✦ Write in Claude” on the sprint to turn it into a running order you can hand to a writer.
One Collection per client. The Sprint list is a single account-level list, so before you start the next client, tick the keywords and hit “Save to collection” in the selection bar, named for the client. Each engagement becomes its own saved list you can reopen from your dashboard any time (it reloads into the tool as a working set with fresh data). Then Clear the sprint and start the next client clean, otherwise the one Sprint list carries everyone.

You end with: a pitch built on their own data and a scoped quarter of work.

What the numbers actually mean

Opp: Is this worth pursuing, demand, commercial value, momentum and how winnable it looks. It is not a difficulty score. Hit Adjust score weights to dial demand, commercial, winnability, momentum, discovery and buyer fit up or down, and the whole table re-ranks live.
Fit (Buyer Fit, Pro): How well a keyword fits the buyer you saved, scored 0 to 100 by your own AI against the buyer profile in your dashboard. Not how hard, not how many, who. Save one profile per thing you promote, pick one in the toolbar, and hit ✦ Score against my buyer. Sort by it, filter to High fit only, or fold it into Opp with the buyer-fit weight.
Ad comp: How hard advertisers bid on it. Proof that a keyword converts, and nothing at all to do with how hard it is to rank. A term with no advertisers can still be impossible.
KD: Real organic difficulty, worked out from the backlink profiles of the pages ranking now. Optional, because it costs a call.
Links to beat: The referring domains those ranking pages actually have. More useful than KD: two keywords a few points apart can need twentyfold different link counts.
Value: Volume × CPC. What that traffic would cost per month if you bought it as ads.
Eff. vol: Volume after a CTR haircut for the blocks above the results, AI Overview, local pack, shopping, People Also Ask. An estimate for comparing keywords fairly, not a traffic forecast. Appears once you've checked page 1.
Open?: How winnable page one is without your own authority. High means UGC platforms rank there and you can compete today.
My pos: Where you already rank, from the Search Console file you uploaded. 2–15 is the fastest money on your whole site.
ON YOUR SITE: You already cover this keyword. Paste your sitemap into Check my site and every result is matched against your published pages, so a search shows what's a genuine gap and what you've written already. Flip to gaps-only to hide what you've covered.
JUMP: Shares no words with your seed. Google connects them by meaning. These are the keywords you didn't know to ask for.

Cost, limits and what happens to your data

It's free and there's no signup. No account, no seat limit, no trial that expires.
Your Search Console file never leaves your browser. It's read locally and joined to the results on your own machine. Nothing is uploaded and nothing is stored on a server.
Your Sprint list is saved in your browser, and synced to your account when you're signed in, so it's still there tomorrow and follows you across devices. It's one list per account, not one per search: to keep separate campaigns apart, save each to a named Collection and clear the sprint between them.
Two actions cost real money to run, “Add SEO difficulty” and “Check page 1”. Both show you the exact cost first, and both have daily caps so one visitor can't run up the bill. Everything you've already looked at is cached and free to look at again.
The data is Google Ads keyword data, resold through DataForSEO. Real volumes, not the bucketed ranges Keyword Planner gives you without ad spend.

I pay for that data out of my own pocket, deliberately, because I'd rather you came back than hit a paywall. If the tool earns its keep, that's the whole business model.

How to actually use this