Faragiri

Learn Finding keywords Keyword research

A North Yorkshire insulation firm built its first list of 214 queries without a keyword tool, from its own inbox, its own reviews and four other free sources.

4Sections
3Min read
4Depth

Start from questions somebody has already asked, not from a tool. Ingleton Insulation installs loft and cavity insulation across North Yorkshire and built its first list in April 2026 without opening anything paid: 214 queries from six sources, and 61 of them from the inbox alone.

The order matters as much as the sources. Harvest first and judge afterwards, in that order, because a list assembled with one eye on whether each row could rank is a list already narrowed by an opinion formed before the evidence. The judging is a separate afternoon.

The six, and what each is good for

The inbox is the highest quality source and the smallest. Six months of enquiries gave 61 distinct questions, in the customer's words, every one of them from somebody who was actually buying. Nothing else on this list has that property, and no tool has it at all.

Sources

Six places a first list comes from, and none of them is a keyword tool

They touch because they overlap: the same question arrives from several places in different words, and the overlap is how you know it matters.

214 queries
  • Your own inbox61 distinct questions in the buyer's words, all from people actually buying.
  • Search suggestions34 phrasings, including one job the firm does and had never written about.
  • People also ask40 follow-up questions that belong as sections of a page rather than separate pages.
  • Rivals' headingsPublished titles and subheadings from competing installers reveal their target list.
  • ReviewsBuyer feedback records the exact problem phrases customers use day to day.
  • Search ConsoleQueries the site already appeared for unnoticed once it has been live long enough.
  • Your own inbox61 distinct questions in the buyer's words, all from people actually buying.
  • Search suggestions34 phrasings, including one job the firm does and had never written about.
  • People also ask40 follow-up questions that belong as sections of a page rather than separate pages.
  • Rivals' headingsPublished titles and subheadings from competing installers reveal their target list.
  • ReviewsBuyer feedback records the exact problem phrases customers use day to day.
  • Search ConsoleQueries the site already appeared for unnoticed once it has been live long enough.

Harvest first and judge afterwards so opinion does not narrow the list before the evidence is in.

Search suggestions give the phrasing, not the demand. Typing loft insulation into a search box and reading what completes it produced 34 phrasings, including loft insulation grant and loft insulation removal, which is a job Ingleton does and had never written a word about. What suggestions cannot tell you is how many people type each one, and treating a long list of suggestions as a list of opportunities is the commonest way this source is misread.

The people also ask box is a list of the next questions. Opening two of them opens four more, and Ingleton stopped at 40. These are useful in a specific way: they are the questions somebody has after the one they typed, which makes them the sections of a page rather than pages of their own.

The three that come from other people

A rival's headings are their keyword list, published. Three competing installers between them had 47 page titles and 130 headings, and pulling those into a sheet took twenty minutes. This is not copying; the words are the evidence of what those firms decided to target, and half of them are wrong, which the next afternoon establishes.

External signals

What came from competitors, buyer reviews, and live impressions

Three outside sources supplied the remaining rows once internal emails and search boxes were exhausted.

  • 130Rival headings from 47 titlesPulled from three competing installers in twenty minutes to see their target list.
  • 22Phrases from 90 reviewsExact buyer wording, including cold bedroom above the garage used five times.
  • 39Search Console queriesTerms the site already triggered without anyone deliberately writing for them.

Half of a rival's target list turns out to be wrong once checked, which is why evaluation happens on a separate afternoon.

Reviews contain the words buyers use for the problem. Ingleton read 90 reviews of competitors and its own and pulled 22 phrases, of which cold bedroom above the garage appeared five times and appears on nobody's site in the county.

Search Console supplied the last group, 39 queries the firm already appeared for without knowing it. That source only exists once a site has been up long enough to appear for anything, which is why it is last here and first in most guides.

Where this stalls: harvesting only what confirms the business

The stall is a list that describes the services page. Everything in it is a variation on loft insulation cost, because that is what the firm sells and therefore what somebody sitting in the office thinks of.

The stall

Office assumptions versus outside sources

Harvesting only from what the team thinks it sells hides real services like loft insulation removal.

Office-only listBuilt from what staff already call the service
Outside-the-building sourcesSuggestions, rival headings, and buyer reviews
Starting pointExisting services page wordingWhat searchers and reviewers actually type
Typical queries foundVariations on loft insulation costUnwritten jobs like loft insulation removal
What gets missedAbout 40 real household jobs a yearNothing hidden by internal jargon

A query nobody ever emailed you directly is still worth inspecting when an external source surfaces it.

A query nobody has ever asked you is worth a second look. Loft insulation removal was on the list because a search box suggested it, not because anyone at Ingleton had thought of it, and it turned out to be a job the firm does for about 40 households a year and had never mentioned anywhere. The sources that produce those are the three that come from outside the building.

What it cost, and what it did not include

The whole exercise was one afternoon and no money. Google Search Console supplied the 39, free, and it replaced the keyword data Google removed from analytics in 2011; the rest came from a browser and a review site. Google Keyword Planner was opened afterwards to attach rough sizes to the 214, free with a Google Ads account, and it reports wide ranges rather than numbers when the account has no spend in it, which is exactly good enough for sorting a list into three piles.

Scope

What a free afternoon buys, and what it leaves undecided

Zero-cost sources and rough Keyword Planner ranges sort 214 queries into piles, but deciding intent and page order comes next.

  • Delivered by the harvestFree browser, reviews, Search Console, and Keyword Planner
    • 39 existing queries from Search Console
    • 214 raw candidate queries in one sheet
    • Wide volume ranges to sort into three piles
  • Still left to decideSeparate judgement after harvesting
    • Whether two queries share one searcher intent
    • Which rows justify effort against competition
    • Which page to write first

One free afternoon

Google Keyword Planner is free with an Ads account and gives wide ranges without spend, enough for rough sorting.

The 214 is a list and nothing more. What each row means, and whether two rows are the same question, is what the searcher wants, which the words hide. How much of the list is worth the trouble is specific queries, less competition, better conversion, and the numbers now attached to each row are two numbers, and how badly both mislead. All of it feeds finding what people type, and what they mean by it, and none of it says which page to write first.

Alongside this

Above this