Keyword Research for Beginners: How UK Small Businesses Find Terms They Can Actually Win

Quick answer

Keyword research is just finding out what your customers type into Google, then being honest about which of those searches you can realistically show up for. Small businesses get it wrong in one predictable way: chasing the two-word trophy terms (“plumber”, “logo design”) that established sites monopolise, while ignoring the longer, specific searches (“boiler making banging noise”, “how much does a shop sign cost”) that have buyers behind them and far less competition. The free tools are enough; the discipline is intent + honesty about difficulty.

Finding what people actually search (free)

• Google’s own suggestions: type your service and let autocomplete finish it; scroll to “People also ask” and “Related searches”. This is Google telling you, verbatim, what gets searched.
• Your enquiry inbox: the questions customers ask by phone and email are the questions others type into Google. A month of enquiries is a content calendar.
• Search Console (if the site has history): the Queries report shows searches you already appear for — terms at positions 8–20 are your ready-made shortlist, because improving an existing foothold is far easier than building a new one.
• Competitor page titles: the businesses ranking where you want to be have done the research; their service page titles are it, in plain sight.
• Free tool tiers (Google Keyword Planner via any Ads account, or the free allowances on mainstream SEO tools) for rough volume — treat the numbers as relative, not precise.

Intent: the sorting that matters more than volume

A 30-search/month buy-now term beats a 3,000-search informational one for a small business, every time. Volume is vanity; intent is revenue. Cost-research terms are the underrated middle — high buying intent, underserved because businesses fear naming prices.

Being honest about difficulty

Search your target term and read page one like an auditor: if it’s national brands, directories and decade-old specialists, a new site won’t crack it this year — that’s an authority contest, and authority takes time (how long, honestly). The winnable ground for a small site: longer phrases (4+ words), question forms, specific pain points, and underserved niches where page one is thin forums and generic content. The compounding strategy is clusters — own twenty specific searches around a topic and you build the authority to compete for the bigger heads later. That’s also why keyword research and owned-audience building pair well: long-tail visitors captured once compound forever.

Mapping: one intent, one page

The output of research isn’t a list — it’s an assignment of terms to pages: each core service search gets a dedicated service page (the architecture argument from one-page vs multi-page), each question cluster gets a post, and closely-related phrasings (“flyer design price” / “flyer design cost”) share one page rather than competing thin duplicates — the cannibalisation trap. Write the page title from the search’s own words, answer the intent in the first paragraph, and you’ve done more than most competitors. The craft of turning a keyword into a page that deserves to rank is its own subject — and if you’d rather buy the whole discipline than learn it, that’s what SEO costs and what our SEO service does.

A worked example: one trade, twenty minutes

A kitchen fitter in Kent, starting from zero: typing “kitchen fitting” into Google autocomplete yields “kitchen fitting cost”, “kitchen fitting near me”, “kitchen fitting course” (skip — wrong intent). People-also-ask adds “How much does it cost to have a kitchen fitted UK?”, “Do kitchen fitters do plumbing?”, “How long does a kitchen fit take?”. The enquiry inbox contributes “can you fit a kitchen we bought from Howdens?” — a question with a buyer attached and almost no competition. Twenty minutes produces a map: a service page targeting “kitchen fitting [town]”, a pricing guide for the cost question, and three posts answering the PAA and Howdens questions, each linking to the service page. That’s a quarter’s content plan — no tools paid for, every topic traceable to a real search. The same exercise works for any trade or service: your market is already telling you what to write; research is just listening systematically.

Read the shape of the results page, not just the competitors on it

Before a term goes on your list, look at what the results page is made of, because the layout tells you how much of the click is even available to a blue link. Four patterns come up constantly for UK small businesses:

• A map pack sitting above everything. For “[service] near me” and most town-plus-trade searches, the first thing on screen is three local listings. Ranking first in the ordinary results below a map pack is worth far less than being in the pack, which makes this a listing job rather than a page job — the work is in getting the free Google Business Profile listing right before you write a word of new content.
• Four ads then a directory wall. Checkatrade, Yell, Trustpilot and the comparison sites occupy the top organic slots on many commercial terms. You can rank fifth and still be invisible. Either target a longer variant, or accept that this particular term is bought rather than earned and treat it as a paid search line item instead.
• A page dominated by questions and explainers. People Also Ask boxes, forum threads and how-to articles mean Google has judged the intent as informational. Put a sales page there and it will not rank however good the offer is.
• A thin page of generic content. Two aggregator pages, an unrelated national brand and a 300-word article from 2019. This is the winnable ground, and it is more common on specific, awkwardly-phrased searches than anyone expects.

Do this check on a clean browser or an incognito window, and if the term has a local element, remember your own location is skewing what you see. The shape of the page is a better difficulty signal for a small site than any single number a tool gives you.

A scoring sheet for turning a long list into a short one

Thirty candidate terms and no method produces paralysis. Score each one out of three across four columns and sort the spreadsheet. Anything scoring 9 or more goes into this quarter’s plan; anything under 6 goes into a “later” tab you revisit next year.

The third column is the one people fudge, and it is the most important. A term you can rank for but cannot back with real work, real photographs and a real price produces traffic that bounces, which teaches Google nothing good about the page. If the honest score is a 1, the answer is not better writing — it is a different term.

Where the free volume numbers mislead

Treat every search volume figure as a rough band rather than a measurement, for four specific reasons. Keyword Planner reports in buckets when an account has little spend, so “10–100” is doing a lot of work. Tools group close variants together, meaning one reported number may be spread across a dozen phrasings you would write separately. Seasonality is flattened into an annual average, so a term that spikes in October and dies in February looks steady all year — worth checking the trend view before planning content around it. And the smallest, most valuable searches often report as zero volume simply because they fall below a reporting threshold, which is exactly why hyper-specific buyer questions look worthless in a tool and convert beautifully in practice.

The reliable numbers come from your own site rather than a tool. Once pages are live, impressions and average position in Search Console tell you what you actually appear for, and your analytics tell you whether those visitors did anything — the small set of figures worth watching is covered in the beginner’s guide to website analytics. Six months of your own data beats any estimate.

The quarterly half-hour that keeps the list alive

Keyword research is not a one-off document. Book thirty minutes at the end of each quarter and do four things. Pull the Search Console queries report for the last three months and look for terms you now appear for that were never on your list — genuine demand you did not predict, and usually the best content ideas you will get all quarter. Check which planned pages actually got written and move the rest forward. Look for pages sitting on positions 8 to 15, because improving one of those beats starting something new. And re-read your enquiry inbox for the phrases customers used this quarter, which change faster than tools update.

Then write the pages properly, because a well-chosen term still needs a page that answers it fully rather than one stuffed with the phrase — the discipline for that sits in writing pages that rank without writing for robots.

Sources & Further Reading

• SEO Starter Guide — Google
• Keyword Planner — Google
• Helpful Content Guidance — Google