Methodology · Visibility
How the Keyword research per product is calculated
Method version visibility-v2 · Last updated
In short
Keyword research per product answers: which Google searches should we rank on to own the search visibility of this product in this country? Claude writes seed searches in the country's languages, DataForSEO expands them into a few hundred real keyword ideas with search volume, difficulty and intent, Claude keeps the ones really about the product, and DataForSEO adds where each tracked brand ranks. Every keyword gets a status (gap, open, improve, owned) and the extra visits you would get at position 3.
Formula
Status: owned = your position ≤ 3 · their brand = a search for a competitor's name · improve = 4–10 · gap = a competitor ≤ 10 and you > 10 · open = no tracked brand ≤ 10 Potential visits = search volume × max(0, CTR(3) − CTR(your position)); CTR(3) = 10%; 0 for a competitor's brand
How it is calculated, step by step
- Claude writes ten short, generic searches per language for the product in the chosen country (no brand names).
- DataForSEO Labs (keyword ideas) expands them into up to 500 related keywords per language (most relevant first) with monthly search volume, keyword difficulty and search intent.
- Claude keeps the keywords someone interested in the product could type and drops other products, other topics and searches for one company's login or website.
- The 150 highest-volume keywords per language are kept.
- DataForSEO Labs (ranked keywords) gives each tracked brand's Google position (top 20) on those keywords.
- Each keyword gets its status and potential visits; the list is sorted by potential.
What goes into it
- Product
- A product line from your market profile, or any product you type.
- Country
- The market and its languages (e.g. Belgium: Dutch and French).
- Tracked brands
- You and the competitors in your workspace.
Data sources
- DataForSEO Labs (keyword ideas, ranked keywords) for Google in the chosen country.
- Claude (Anthropic) for the seed searches and for selecting the keywords about the product.
How often it is updated
On demand, up to 20 researches per organization per month. Results are kept.
How to read it
- Gap keywords with low difficulty are the quickest wins: a competitor proves Google wants such a page and it is not hard to rank.
- Open keywords are usually won by comparison sites, media or forums; content that answers the search well can win them.
- Commercial and transactional intent are closest to a sale; informational searches build reach earlier in the decision.
Example
Mortgage loans in Belgium: a search with 6,600 searches a month and a difficulty of 25, where a competitor ranks 2nd and you are not in the top 20, is a gap; reaching position 3 would add about 6,600 × 10% = 660 visits per month.
Limitations
- Keyword ideas come from DataForSEO's database; very new or rare searches can be missing.
- The selection by an AI model can occasionally keep or drop a borderline keyword.
- Positions are from DataForSEO's latest database update, not a live check of the results page.
Frequently asked questions
How is this different from the keyword gap?
The keyword gap only looks at keywords the tracked brands already rank for. Keyword research starts from the product and the country, so it also finds searches no tracked brand ranks on yet.
What does a research cost?
It is included in your plan up to 20 per month; behind the scenes it uses paid DataForSEO and Claude calls.
Related
- Keyword gap — Keywords where a competitor ranks in Google's top 10 and you don't, sorted by the extra visits per month you would get at the best competitor's position.
- SEO per product cluster — Every search keyword is sorted into one of your product lines, General or Not your market; share of voice is then calculated per product, and Not your market doesn't count.
- Keyword difficulty (KD) — How hard it is to reach Google's top 10 for a keyword, from 0 to 100: easy below 30, medium 30–59, hard 60 and above.
- Estimated search traffic — Estimated monthly visits from Google per keyword: search volume × the click-through rate of the brand's position; summed per brand.
