I Stopped Juggling Three Builds. KeywordIntent Is What One Build at a Time Produced.

Build in Public · #18

For several days I worked on one tool and nothing else. KeywordIntent went from a very good free keyword tool to a Pro version with features I have not seen combined in any keyword tool. The biggest reason was not the code. It was testing it between every update.

KeywordIntent started with one hook: search “running shoes” and it finds “asics gel 1130”, keywords that share no words with your seed at all. That still sits at the heart of it. What changed in this sprint is everything built around it, and the way it got built.

Why one build at a time changed the result

I usually have three builds on the go at once. Focus gets split three ways and none of them gets real attention. This time I picked one tool and gave it several days of full focus, with a single goal: make it as good as it can possibly be, and push the boundary of what could be built into it.

The habit that made the difference was testing between updates. Not a quick click around, but using it for real on real keyword research. Almost every test session showed me the next thing to build, and most of the features below came out of using the tool rather than planning it.

Scoring keywords by who is searching

Most keyword tools tell you how many people search and how hard a term is to rank for. Buyer Fit answers a different question: who is searching, and are they your buyer?

You save named buyer profiles: what you sell, who it is for, and the pain it solves. Then you click “Score against my buyer” and every keyword gets a Fit score from 0 to 100, worked out by your own AI. You can sort by fit, filter to “High fit only”, fold fit into the Opportunity score with a weight slider, and pull a “Best 20 for my buyer” shortlist.

It runs on your own AI key, so there is no per-use charge from me for any of it.

KGR+, a whole new-site workflow instead of one number

The Keyword Golden Ratio is a well known way to find terms a new site can actually rank for. I did not just add it as a metric. I built a full workflow around it.

• Add KGR fetches the allintitle count for each keyword and scores the ratio, in a colour-coded column sorted golden-first, with a “KGR green only” filter.
• Classic or Custom thresholds, so you can use the textbook rules or set your own.
• An optional allinurl check as a second signal.
• An article outliner that clusters your golden keywords. You add context about your offer by pasting sales copy, a URL or a transcript, or by attaching a Buyer Fit profile, and your AI drafts one outline that covers the whole cluster. “Write the full article in Claude” hands over a complete prompt.
• A one-click New Site Playbook preset, with a guided walkthrough of the whole flow.
• Export for Site Sniper, my companion tool, so golden keywords go straight into building a microsite.

The outliner is the part I am most pleased with. Classic KGR points you at one page per keyword. KGR+ lets one article target a whole cluster of golden keywords.

A guide that points at the button you need

Powerful tools have a problem: the first time you open one, you do not know where to start. So I built Guide me.

Pick one of 13 real strategies and a coach panel walks you through it step by step, highlighting the exact control you need for each step, right there on the page. The strategies cover ecommerce, local, Google Business Profile, local lead gen, niche and authority sites, blogs, affiliates, SaaS, AEO and AI search, info products and courses, coaches and consultants, agencies, and the New Site Playbook.

I do not know of any other keyword tool that does this. It exists to stop that first-run overwhelm.

The rest of what shipped

• Bulk competitor miner. Paste a list of rivals and one action pulls every keyword each brand appears in, merged into one table and tagged with which rival surfaced it.
• Adjustable Opportunity weights. Sliders to tune demand, commercial value, winnability, momentum, discovery and buyer fit.
• Action Plan with buyer stages. A plain-language brief on what to do with the results, split into bottom, middle and top of funnel.
• Content gap against your sitemap. Paste your sitemap and every result is flagged as already on your site or a genuine gap.
• Similar-intent seed expansion. Your AI turns a thin seed into same-intent terms worth researching.
• AI Overview analysis. Flags keywords that trigger an AI Overview, reads the AI answer and its sources, and shows where you could get cited.
• Smarter presets. Save a whole setup, filters, sort and all, and reload it in one click.

That sits on top of the previous sprint, which added competitor domain analysis with weak-spot filters, organic competitors and a keyword gap, page-level weakness scoring, and renameable history with Collections.

What testing changed along the way

This is the part a feature list never shows. Each of these came from using the tool, not from a plan:

• ICP became Buyer Fit. I realised “ICP” was too close to the name of an existing tool, so I renamed the whole feature to something that is clearly its own.
• Custom thresholds exist because of real volumes. KeywordIntent’s volumes come from Google Ads data, which runs much higher than many tools. The textbook “under 250 searches” rule would wrongly throw out terms that are genuinely open. Custom thresholds fix that.
• The golden-first sort was wrong at first. Sorting by the raw ratio floated excluded high-volume terms to the top instead of the real green winners. It now ranks by verdict.
• The walkthroughs are generated from the guide. Guide me reads its steps live from the how-to page, so the two can never drift apart.
• I got stuck in my own walkthrough. Testing the SaaS route, I hit a vague step. So Guide me moved to a prominent button in the top bar, every step now names its exact location, and each one highlights its control on the page.
• Bulk mining turned out to be cheap. I realised both halves of a brand search come from one lookup, so mining ten rivals takes ten calls, not twenty. That is what made the bulk miner worth building.
• The Site Sniper link needed no extra work. Site Sniper already imported the exact shape KeywordIntent exports, so the two tools connect with nothing to change on either side.

What it costs, and what comes next

The free version stays free, and it got more powerful in this sprint too.

Pro will come in two tiers: Pro Core at $29 a month and PRO+ at $49 a month. You bring your own AI key and your own DataForSEO key, so you are never paying me a markup every time you run a search.

Tomorrow I start on the next stage: automating the workflows, so KeywordIntent runs more like an agent that does the steps for you rather than a tool you drive. If I can pull that off, it becomes a third tier, KeywordIntent Automation at $67 a month. That one is not built yet, so treat it as the plan rather than a promise.

My Take: For a long time I ran three builds at once, and none of them got the attention it deserved. This time I picked one tool and gave it several days of full focus, and the difference was bigger than I expected. Most of it came from one habit: testing KeywordIntent for real between every update. Each test session showed me the next thing to build, things I would never have spotted from a feature list. Buyer Fit, the Guide me walkthroughs and the KGR+ outliner all came out of using it, not planning it. The goal was simple: make this tool as good as it can possibly be and see how far it could be pushed. That focus is what produced features I have not seen combined in one keyword tool, and it is how I plan to build from here.