Keyword research done badly is why so many South African websites rank for terms nobody searches and miss the ones that actually pay the bills. This guide walks through the exact process for finding, filtering, and prioritising keywords for a South African audience in 2026, without relying on generic global tools that assume everyone searches like an American.
TL;DR
Keyword research for South African SEO means filtering global keyword data through local search behaviour: Rand-denominated pricing terms, city-specific modifiers (Joburg, Cape Town, PE, not just "South Africa"), and load shedding-era search patterns that spike around service disruptions. The term "keyword research for south african seo" itself only pulls around 180 monthly searches with a difficulty score of 29, which tells you the real opportunity in this space sits in the long-tail, industry-specific variations, not the head term. Verdict: build your keyword list from Google Search Console data first, then layer in tool-based volume estimates second. A South African e-commerce store or law firm chasing generic keyword volume from a US-calibrated tool will waste months on terms with zero local intent.
Why this matters
Most South African businesses run keyword research through Ahrefs or Semrush, take the volume numbers at face value, and build content plans around search terms that barely register locally. The problem is that most keyword tools default their volume estimates to global or US data, then apply a rough multiplier for South Africa. That multiplier is often wrong by 30% to 50% in either direction, especially for city-level and industry-specific terms.
The fix isn't a better tool. It's a better process that cross-checks tool data against what South Africans are actually typing into Google, which you can only get from Search Console, competitor SERPs, and direct observation of local buying language. Get this step wrong and every downstream SEO decision, content brief, page structure, is built on sand.
What you'll need
- Access to Google Search Console for your domain (or a domain in the same category if you're starting from zero)
- A keyword research tool with South Africa-specific filtering: Ahrefs, Semrush, or Google Keyword Planner set to South Africa geography and ZAR currency
- A spreadsheet (Google Sheets works fine, no need for anything fancier)
- 3 to 5 hours of uninterrupted time for the first pass
- Familiarity with your own service area: which cities you serve, what you charge, and who your real competitors are
The steps
1. Pull your existing Search Console data first
Before touching any third-party tool, export the Queries report from Search Console for the last 16 months. This shows you what Google already thinks your site is relevant for, including terms you're ranking for on page 2 or 3 that you didn't deliberately target.
This matters because it's real click and impression data from South African searchers, not a modelled estimate. A query showing 40 impressions and a 3.2% click-through rate over three months is worth more than a tool's guess at monthly volume.
Common mistake: ignoring queries with low impressions. Some of the highest-converting terms in South African SEO have volumes under 50 a month but convert at 8% or higher because the searcher is close to a buying decision.
2. Build your seed list from real business language, not tool suggestions
Write down the exact phrases your customers use when they call or email you, not the phrases you assume they'd Google. A Johannesburg-based medical practice might get calls asking about "gap cover specialists near me" long before a keyword tool surfaces that phrase as a suggestion.
Cross-reference this against how competitors phrase their own service pages. If three competing law firms in Sandton all use "commercial litigation attorney Johannesburg" as a page title, that's a strong signal of established local search intent, regardless of what the tool's volume column says.
Common mistake: relying only on tool-generated "keyword ideas" that surface generic global variations irrelevant to South African phrasing.
3. Layer in tool-based volume and difficulty, filtered to South Africa
Run your seed list through Ahrefs or Semrush with the location filter set specifically to South Africa, not "Africa" or left on global default. Record volume, keyword difficulty, and the top three ranking domains for each term.
As a benchmark, a term like "keyword research for south african seo" sits at roughly 180 searches a month with a difficulty of 29, which is a realistic, winnable target for a mid-authority domain in 2026. Compare every keyword on your list against that benchmark: anything under difficulty 35 with local competitors ranking is worth prioritising early.
Common mistake: chasing keywords with difficulty scores above 50 in year one. That budget is better spent on 10 winnable long-tail terms than one competitive head term.
4. Segment by city and by buying stage
Split your keyword list into two columns: geography (national, Johannesburg, Cape Town, Pretoria, Durban) and intent stage (awareness, comparison, ready-to-buy). A term like "seo agency" is national and top-of-funnel. "best seo agencies in cape town" is city-specific and comparison-stage, which usually converts far better because the searcher is actively vetting vendors.
This segmentation is what separates programmatic South African SEO from generic content. It's also why Aion Marketing structures client keyword maps by city and vertical before writing a single page.
Common mistake: treating "South Africa" as a single search market. Cape Town, Johannesburg, and Durban searchers behave differently, and Google's local pack results reflect that.
5. Check search intent against the actual SERP, not the keyword's category
For every keyword on your shortlist, open the live SERP and look at what's actually ranking: articles, product pages, local packs, or directories. A keyword tool might label a term "informational," but if the top five results are all service pages, the real intent is commercial.
This is the single biggest reason keyword research fails: matching the wrong page type to the intent Google has already decided the term deserves. A 2000-word blog post won't outrank five local business listings for a term where the SERP shows a map pack.
Common mistake: writing a blog post for a keyword where the SERP is dominated by service or product pages.
6. Prioritise by a simple ratio: opportunity divided by difficulty
Once you have volume, difficulty, and intent confirmed for each term, rank your list using a rough opportunity score: monthly volume divided by keyword difficulty. A keyword with 150 volume and difficulty 20 scores higher (7.5) than one with 500 volume and difficulty 60 (8.3 is close, so check competitor domain strength too before deciding).
This single calculation stops teams from spending three months chasing a head term with difficulty 55 while ten winnable long-tail terms sit untouched.
Common mistake: prioritising by volume alone. Difficulty and your domain's current authority matter just as much.
7. Map each keyword to a specific page, not a general topic
Every keyword on your final list needs one dedicated page. If you're building out service pages for South African e-commerce clients, for example, map "seo for online stores south africa" directly to a page built for that vertical, not a generic services page trying to rank for everything.
This is where a lot of South African businesses lose the plot: they build one broad "SEO services" page and expect it to rank for 40 different keyword variations. It won't.
Common mistake: mapping five keywords to one page and diluting relevance for all of them.
Troubleshooting
- Tool volume numbers look too high or too low compared to Search Console data. Trust Search Console for terms you already rank for. Tools estimate; Search Console reports what actually happened.
- Every keyword idea the tool suggests looks the same as your competitors' pages. That's expected. Differentiate through the specificity of your seed list (step 2), not by hoping the tool finds something unique.
- Difficulty scores don't match what you see in the actual SERP. Difficulty scores are backlink-based estimates. Check the top three ranking domains manually. If they're all low-authority local sites, the term is more winnable than the score suggests.
- You're getting impressions in Search Console but almost no clicks. This usually means your title tag doesn't match search intent, not that the keyword is wrong. Rewrite the title before abandoning the term.
- City-specific keywords show almost no volume in the tool. This is common for smaller South African metros. Use Search Console and Google Trends' relative interest data instead of absolute volume for towns outside the six major metros.
- You keep finding the same 20 keywords no matter which tool you use. Widen your seed list using actual customer language (step 2) and competitor SERP scraping rather than tool-generated suggestions alone.
Tools and resources
- Google Search Console (free, and the most accurate South African data you'll get)
- Google Keyword Planner set to South Africa and ZAR
- Ahrefs or Semrush with location filtering applied
- A spreadsheet template splitting keywords by city, intent stage, and opportunity score
- Review how South African agencies structure their own keyword-mapped pages: see how Aion Marketing's SEO agency directory segments by city and vertical
- If you're in e-commerce specifically, the keyword clusters differ enough that it's worth reviewing a dedicated approach to SEO for online stores
What to do next
Once your keyword map is built, the next bottleneck is usually content structure and technical execution, not more keyword research. If you're running paid alongside organic (most South African businesses should, since SEO takes 6 to 12 months to show results), check how Google Ads copy for e-commerce uses the same keyword data to shortcut paid testing before committing to long-form content.
FAQ
What's the best free tool for South African keyword research?
Google Search Console is the most accurate free source because it reflects real South African search behaviour on your domain, not modelled estimates. Google Keyword Planner (set to South Africa geography) is the best free option for new keyword discovery.
Is Ahrefs better than Semrush for South African SEO?
Both tools apply similar modelling for South African volume data, and neither is dramatically more accurate than the other locally. The bigger factor is whether you filter results to South Africa geography specifically rather than global or "Africa" defaults.
How much search volume is enough to target a keyword in South Africa?
There's no fixed threshold. A term with 50 monthly searches and difficulty under 20 can be worth more than one with 500 searches and difficulty 60, especially if the lower-volume term converts at a higher rate.
Do city-specific keywords (Johannesburg, Cape Town, Durban) matter more than national ones?
For local service businesses, yes. City-specific terms show clearer commercial intent and lower competition than broad national terms, and they match how South Africans actually search when comparing local vendors.
How long does keyword research take for a new South African website?
A thorough first pass takes 3 to 5 hours for a focused service business, longer for e-commerce sites with hundreds of product categories. Budget a half-day minimum before starting content planning.
Should I target keywords in English only, or include Afrikaans and other local languages?
It depends on your audience. Businesses serving Afrikaans-speaking regions (parts of the Western Cape, Free State) often see meaningful search volume in Afrikaans variants that English-only keyword research misses entirely.
What's the difference between keyword difficulty and actual ranking difficulty?
Keyword difficulty scores estimate based on backlink profiles of ranking pages. Actual ranking difficulty also depends on content quality, page experience, and how well your site matches search intent, none of which the score captures.
Is keyword research a one-time task or ongoing?
Ongoing. Search behaviour shifts with seasons, economic conditions, and platform changes. Revisit your keyword map every 6 months, or sooner if you notice ranking or traffic drops in Search Console.
One last thing
The most overlooked keyword source in South African SEO isn't a tool at all, it's your own site search data and sales team call notes. Businesses that cross-reference Search Console queries against what their sales team hears on actual calls consistently find 10 to 15 high-intent keyword variations that never show up in Ahrefs or Semrush, because the phrasing is too specific to South African buying language for any global tool to model accurately in 2026.
