Volume is vanity in Edinburgh: how to research keywords by real revenue instead

Why raw search volume misleads Edinburgh marketers

Every legacy keyword tool leads with the same figure: monthly search volume. On any considered reading it is the most misleading metric in SEO, and nowhere more so than in a capital city where tourism, university term dates and festival scheduling all tug at the same handful of head terms. Take “restaurants Edinburgh” — Ahrefs, Semrush and Keyword Planner all settle on roughly 27,100 searches a month, and it looks, on the face of it, like the jewel of the local hospitality space. In practice, most of that volume is undecided July browsing by inbound tourists, Easter-week research for family day trips, and idle scrolling that never lands in a booking at any particular room. Compare it now with “private dining Marchmont for eight” — seventy searches a month, a group already assembled, a date pencilled in, a spend range mutually understood. One term is worth roughly £9,200 in monthly pipeline to whichever venue ranks first; the other is worth £180. Get the choice wrong, invest six patient months in it, and no finance director in Bruntsfield or the New Town will sign off the next quarter's marketing budget.

Compare that with “private dining Edinburgh New Town for 12” — 90 searches a month. It looks trivial. But every one of those searchers has a group, a date and a spend range in mind, and a strong CTR to whichever restaurant has the right landing page. In our data, that single term is worth roughly £9,400 in pipeline per month for the venue that ranks first. The head keyword is worth £180. Choose the wrong keyword, spend six months on it, and the finance director will not renew the marketing budget.

How the Semalt AI scorer works

Semalt’s keyword engine builds a per-query score on four inputs, blended in a way that reflects how Google actually distributes clicks in 2026:

1. Real click volume — not Keyword Planner ranges, but merged Search Console + clickstream data for UK devices, resolved by device and by day of week.
2. SERP feature layout — live snapshot of what surrounds the ten organic results: AI Overview, PAA, video carousel, Google Business Profile, Reddit thread, image pack. Each feature has its own CTR curve.
3. Intent class — classified by a fine-tuned transformer trained on the last twelve months of UK SERP behaviour, with a specific pass over Scottish tourism, hospitality and public-service verticals.
4. Expected pipeline — clicks × probability of conversion (from category benchmarks or the client’s own historical data) × GBP value per conversion. Sortable, filterable, exportable.

Festival seasonality: Fringe, EIFF, TIFF, Hogmanay

Edinburgh keyword volumes are the most seasonally volatile of any UK city outside London. From late July to the end of August, “Fringe”-modified queries add a 41× multiplier over their October baseline. TIFF (Edinburgh Traditional Fringe) and the International Book Festival stack on top. EIFF (the Edinburgh International Film Festival) drives its own spike in mid-August. Hogmanay queries take over in late December. A tool that reports a single monthly average smooths those spikes into invisibility — and Edinburgh hospitality clients then miss the six-week window that pays for the year.

The Semalt seasonality view shows each keyword’s weekly forecast for the year ahead, with the option to overlay the actual event calendar (Fringe programme dates, Six Nations fixtures at Murrayfield, cruise-ship arrivals at Leith and Rosyth, RBS AGM week for corporate hospitality). Marketers use it to commission content eight to twelve weeks ahead of the spike, which is roughly the time Google needs to index and stabilise a new URL in a competitive tourism SERP.

Scottish English, Scots dialect and tourism phrasing

Most keyword tools were built on US English corpora and treat British spellings as afterthoughts. Semalt’s UK index is native: “jewellers” and “jewelers” are separate entries, “colour” and “color” likewise, and the model knows that in Edinburgh “wee” is a legitimate modifier (“wee bar Rose Street” — 210 searches a month) rather than a typo. A handful of Scots dialect terms turn up in real queries too — “bothy” for a small hut, “stovies” for the traditional dish, “ceilidh” for a dance — and the model treats each as a first-class query rather than a suggested correction.

The lexicon also contains the specific way Edinburgh users refer to their neighbourhoods — “Bruntsfield” rather than “south side”, “the Colonies” rather than “Stockbridge cottages”, “the Meadows” rather than a formal park name. Miss those and your landing page ranks for a synonym nobody uses.

Case study: Royal Mile whisky specialist

An independent whisky retailer on the Royal Mile, with a shop, a tasting room and a fast-growing mail-order line into Germany, Japan and the US, asked us in January 2026 to prioritise their SEO budget. Their previous consultant had them chasing the head term “whisky Edinburgh” (12,100 searches/month) for eighteen months. They were ranked position three and getting decent traffic, but the mail-order sales attributable to that keyword were negligible.

We ran the Semalt AI scorer across their category and the top ten keywords by expected GBP pipeline looked nothing like the traffic-first shortlist:

We commissioned seven landing pages and refreshed three category pages against those queries, then paired the work with Semalt’s AutoSEO to keep on-page signals current as SERPs shifted. Five months in, mail-order revenue was up 52% year-on-year and the corporate-hospitality line — a niche the founder had been considering closing — was booking eleven groups a month, most of them from a single 300-word landing page that ranked first for the corporate query. Log in to Semalt and run the same scorer on your own category.

The recommended workflow for an Edinburgh SME

Semalt vs. Ahrefs vs. Semrush for UK research

FAQ

Does the model understand Scots or only Scottish English?

Both. The lexicon includes the roughly 400 Scots and Scottish English terms that appear in real UK search queries with meaningful volume — ceilidh, bothy, stovies, wee, aye, quaich, dram — and the classifier resolves each to the correct intent bucket rather than treating them as spelling errors.

How current is the volume data?

Weekly refresh for the UK top million queries, monthly for the long tail. During Fringe and Hogmanay windows we tighten to daily refresh on the affected clusters so seasonality captures the actual spike, not the historical average.

Can I import my own keyword list?

Yes — upload a CSV or paste a list, and Semalt returns intent, difficulty, live SERP layout and pipeline forecast per keyword. Useful for legacy PPC keyword sets that have been sitting in a shared drive for two years.

Does the tool export to a content brief?

Yes. Every cluster produces a brief covering title, meta, H1, H2s, target entities, questions to answer, competitor URLs to beat and expected pipeline. Formats: Google Docs, Notion, plain markdown, Contentful JSON.

How does it handle bilingual sites?

If your site ships English plus Gaelic or French for tourism, the engine indexes each variant separately and cross-references hreflang with the technical audit so you don’t accidentally cannibalise your own rankings.

What’s the free tier limit?

Enough to run keyword research on a single domain and export up to 500 keywords a month. Paid tiers unlock competitor comparison and unlimited exports — but for an Edinburgh SME the free tier is often enough to commission the first content sprint. Start on the free tier here.