Cardiff & Swansea: Welsh-market keyword research patterns

Welsh local SEO sits in an odd position. Cardiff (365,000 residents, the political and media capital), Swansea (245,000, the West Wales economic centre), and Newport (160,000) collectively form a market of around three-quarters of a million people — comparable to a single mid-tier English city. But the Welsh competitive ecosystem differs from English markets in three ways that matter for keyword research: a distinct directory and competitor set, bilingual considerations Google handles inconsistently, and a publicly-funded layer (Welsh Government, BBC Cymru Wales, S4C) that locks up specific SERP categories.

Most UK-wide SEO agencies treat Wales as "England, but smaller". The data says otherwise.

Cardiff: government, media, and an unusually contestable middle

Cardiff's SERPs sort into three groups:

Locked. Government services, anything Welsh-Government adjacent, anything rugby-related, and most "things to do" tourism queries. The official .gov.uk, .gov.wales, Welsh Government, Visit Wales, and Wales Online sites have been online for 20+ years with DS 60–90. Don't try.

Contestable. Cardiff residential and neighbourhood queries — Cathays, Roath, Pontcanna, Canton, Whitchurch, Penylan, Llandaff, Cyncoed. Average DS 30–40 in the top 10 with frequent thin content at rank 4–10.

Open. B2B and specialist services. "Web designer Cardiff", "GDPR consultant Cardiff", "PPC agency Cardiff" return top 10s where the average DS is in the high 20s and the rank-1 result is often a Yell or Bark page.

The structural reason for the contestable-and-open layers: Cardiff has fewer well-resourced local agencies per capita than Bristol or Birmingham, so the SEO baseline for local businesses is lower.

Swansea: dramatically softer than Cardiff

Swansea's SERPs are softer than Cardiff's across almost every category. Three factors:

Fewer well-resourced competitors. Swansea has roughly half Cardiff's local agency density and a much smaller pool of mid-sized professional services firms running their own SEO.

Tourism queries are less locked. Visit Swansea Bay and the major Welsh tourism portals exist but don't dominate as completely as Visit Cardiff or Wales Online do. Restaurant and "things to do" SERPs in Swansea have a softer top 10 than Cardiff's equivalents.

Mumbles and Gower position uniquely. Mumbles ("the Mumbles", strictly) and the Gower Peninsula attract genuine tourism intent with surprisingly weak top 10s for service businesses. "Restaurant Mumbles", "estate agent Gower", "wedding venue Gower" return SERPs that don't match the visitor volume.

The bilingual question

Welsh-language search volume is small but the SERP behaviour around it is interesting. Google generally returns English-language results for Welsh-language queries unless the page is clearly Welsh-only — and pages with a Welsh-language variant (with proper hreflang markup) can pick up Welsh-keyed queries with very little organic competition.

The practical play: a Welsh-language landing page for service businesses operating bilingually is genuinely free traffic. Search volumes are small (typically under 500/month for service-category Welsh queries) but the SERP is empty enough that ranking is straightforward.

Keyword patterns that work

[service] + Cardiff suburb — Pontcanna, Canton, Roath, Penylan, Whitchurch, Llandaff. The suburb-keyed queries are routinely 30–40% the volume of the city-keyed queries with one-third the SERP density.

[service] + Swansea / Mumbles / Gower — three positioning angles, three different SERPs. Mumbles and Gower searches over-index on lifestyle, accommodation, and tourism-adjacent services.

[service] + Newport — Newport is often forgotten in Welsh SEO work, but its SERPs are softer than both Cardiff and Swansea. Newport-keyed queries for trades, professional services, and healthcare return top 10s with average DS in the 25–30 band.

Welsh-Government-adjacent B2B — recruitment, consultancy, training, IT services that genuinely serve the public sector in Wales. These return SERPs where the dominant competitor is a London-headquartered firm with weak Welsh-specific content; a clearly-Welsh-positioned page lands surprisingly fast.

Where SERPTool flags consistently

• Aggregator dominance at rank 1–3 with Yell, Three Best Rated, and Welsh-specific Cariad reviews appearing in the top 3 for many service categories.
• Outdated content in the 4–10 band — many Welsh local-business pages haven't been refreshed in 2–3 years.
• Missing Welsh-language variants — even bilingual local businesses frequently don't have a Welsh page, leaving the Welsh-keyed SERP genuinely empty.

Verticals where the gap is widest

• Trades in Newport — the softest local-trades SERPs in Welsh markets.
• Tourism services on the Gower — wedding venues, accommodation, activities, food and drink.
• Specialist B2B in Cardiff — anything outside government-adjacent and finance.
• Health and wellness in Cardiff suburbs — Pontcanna and Penylan especially.
• Welsh-language content — for any service business with Welsh-speaking staff who can sign off on the copy.

The Welsh-market strategy in one paragraph

For Cardiff service businesses, the keyword research output is a Cardiff page, 4–6 suburb pages, and a Newport page if the business genuinely covers South-East Wales. For Swansea, a Swansea page plus Mumbles and Gower variants where applicable. For any bilingual business, a Welsh-language landing page indexed alongside English. Run bulk keyword research across all three cities together — the cross-Welsh-market view often surfaces opportunities a single-city audit misses.