AEO South Africa. Answer Engine Optimization for South African brands.
AEO isn’t a rebrand of SEO. SEO is about beating the algorithm that ranks ten links. AEO is about being one of the three brands an AI engine names when a CMO at Discovery asks ChatGPT for the best medical aid in South Africa. Different game. Different scoreboard.
This is the category page for the discipline. What AEO is, how it differs from SEO, the four engines that matter, how to measure it, why the South African version isn’t the same as the US one, and the action ladder Capitec, Showmax and OUTsurance are already climbing. Read it like a manual.
SA-AEO-Bench v1.2 — pre-registered on OSF before data collection. 19,020 successful AI responses across ChatGPT (GPT-5), Claude Sonnet 4.5, Gemini 2.5 Pro and Perplexity. 232,254 cited URLs classified against a published SA source taxonomy. 100 South African brands. 12 industries. Re-run quarterly. Methodology and analysis code public at osf.io/w4az2.
The per-brand cut is in the Scorecard. The cross-industry cut is in the Q2 Index Report. The methodology lives at /methodology.
What AEO actually is.
AEO is the work of getting your brand named inside an AI answer. Not the seventh result on page one. The answer.
When a South African opens ChatGPT and asks “what’s the best medical aid in South Africa?”, the engine composes a short response and names two or three schemes inside it. Discovery Health. Bonitas. Maybe Momentum. That shortlist is the answer. Every brand outside it is not in the consideration set. They are not losing — they are not in the room.
AEO is what you do to be one of the names inside that answer. The discipline covers what AI engines pull from (citations), what they reason against (entities, sources, editorial signal), and what they output (the named brands). If SEO was about beating the algorithm, AEO is about being the answer the algorithm composes.
The short definition: AEO is the discipline of showing up in AI answers about your category. The longer one: AEO is the practice of shaping the signal AI engines use — citations, entities, sentiment, freshness, source authority — so the engines reach for your brand when a real customer asks a real question.
A few things are not AEO. Schema markup on its own is not AEO — schema helps but a perfectly marked-up page on a quiet domain stays invisible. Generating AI content is not AEO — that’s a content tactic, not a visibility one. Building a chatbot is not AEO. AEO is specifically about what frontier engines say when humans ask them questions.
For a deeper definition pass, the what-is-AEO post walks through it line-by-line. The Cited Brands Glossary gives you the rest of the vocabulary in one page.
AEO vs SEO — the breakdown.
People keep asking if AEO replaces SEO. It doesn’t. The two disciplines sit next to each other and share a supply chain.
Here’s the honest map. Both want your domain to be authoritative. Both want clean structured data. Both want fresh, deep content. From there they part ways — sometimes politely, sometimes hard.
| Dimension | SEO | AEO |
|---|---|---|
| What you optimise | Pages — title, content, links, schema | Brand mentions and citation footprint |
| Where you show up in | A ranked list of ten blue links | A composed answer naming two to four brands |
| The target | Keywords | Questions |
| The win condition | Position 1–3 on the SERP | Named inside the answer at all |
| How you measure | Rank, CTR, impressions, sessions | Mention rate, citation share, narrative control |
| Time horizon | Months to years | Weeks to quarters |
| Source of truth | Google Search Console | SA-AEO-Bench and per-engine probes |
| The hardest part | Out-ranking competitors | Being the name AI volunteers |
| Skills it needs | Technical SEO, content strategy | Editorial PR, entity work, AI eval design |
| Who owns it | Marketing or growth | Marketing, PR, brand and CX together |
Where they overlap: ranking matters to AEO. AI engines reach for the same authoritative sources Google reaches for, so an article that Google ranks page-one for “best bank South Africa” is also one Claude will lean on. The citation core that wins SEO partly wins AEO too. Your technical SEO floor is the floor underneath the AEO programme.
Where they conflict: SEO optimises for the click; AEO optimises for the mention. AI answers reduce clicks. A ChatGPT user who gets “Discovery, Capitec and FNB are strong choices” in an answer may never click through to any of those three sites. The brand still won — they got named — but the SEO dashboard sees flat sessions. If you run the two programmes on separate scorecards, they look like they’re fighting. They’re not. They’re measuring different parts of the same funnel.
Honest take: most of what your SEO team already does still applies. The difference is the goal. SEO chases rank; AEO chases being named. The work product on a strong AEO programme looks like editorial PR and entity engineering with an SEO foundation underneath — not SEO with an AI bolt-on. The free Scorecard makes the gap concrete for your brand: where you rank well but get ignored by AI, and where AI names you despite a middling SEO position.
The four engines that decide.
Four engines do almost all the work today. They answer differently. They cite differently. They forget differently. A single AEO play almost never works across all four.
ChatGPT is the most-used answer engine for SA high-consideration decisions — medical aid, banking, car-buying. Its retrieval leans on durable editorial coverage: BusinessTech, Moneyweb, TechCentral, Wikipedia. It tilts toward the brands with the longest reputational tail. Capitec, Discovery, MTN, Showmax all win here.
What SA brands need to know: ChatGPT rewards depth and longevity, not freshness. A six-month editorial programme on the right SA sites lifts you here. A one-off press release does not.
Claude is the editorial engine. It privileges sources that explain themselves — research bodies, regulators, peer-reviewed work, named bylines. It is the most likely to surface a caveat in the same paragraph as a recommendation. For SA brands, Claude rewards methodology pages, transparent pricing, regulator relationships and the kind of editorial coverage Moneyweb and Daily Investor produce.
Bias to know: Claude over-weights international benchmarks unless an SA editorial source explicitly contextualises the local market. That gap is where SA challenger brands get under-cited.
Gemini is the consumer-facing engine. It powers Google AI Overviews, which is what most South Africans actually see when they search. Gemini cites Reddit hundreds of times on SA queries; ChatGPT and Claude cite it zero times. It also reaches deep into Hellopeter, MyBroadband forums and consumer-review aggregators that the other two engines effectively ignore.
If your audience starts a search on Google, this is the engine that frames the answer. SA brands with active Hellopeter complaint volume feel it here first.
Perplexity is the citation-first engine. It cites inline, prefers recency, and treats freshness as a primary ranking signal. A SA brand that has been in the BusinessTech or TechCentral cycle in the last fortnight reads as the current authority here. A brand that has gone quiet drops out fastest.
Perplexity is the smallest by audience but the most influential by analyst-and-journalist habit. It is where competitive intelligence is built before it lands in CMO decks.
Four engines, four reading patterns. A brand that owns ChatGPT can be invisible in Gemini. A brand killing it on Perplexity may have decayed out of Claude entirely. The per-engine spread for your brand is in the Scorecard and the cross-engine asymmetry by SA industry is in the Q2 Index Report.
What to actually track.
AEO measurement is its own discipline because the SEO dashboard cannot see most of what matters here. Five things. The rest is noise.
Does AI name you at all?
Definition: across an organic, unprompted set of category questions, the percentage of answers that name your brand at all. Run blind — no brand name in the prompt. This is the most honest single number in AEO. It tells you whether you are in the conversation or outside it.
Why it matters: every percentage point of organic mention is direct AI-channel demand capture. Cited Brands measures it per engine on every brand in the bench, with blind and named-prompt versions so we can separate real visibility from prompt-cue inflation.
Whose pages does AI reach for when it answers about you?
Definition: of the URLs cited inside answers about your category, the share that points to sources friendly to you — your own domain, editorial coverage that frames you well, regulator pages, association content. The rest goes to complaint platforms, competitor comparisons or international sources.
Why it matters: mentions get you into the answer; citations decide what the answer says about you. Citation share is where reputation lives.
Own. Editorial. Complaint. What’s the mix?
Definition: the breakdown of citations about your brand into three buckets. Own-channel (your domain, your documentation, your press releases). Editorial (third-party journalism — BusinessTech, Moneyweb, MyBroadband). Complaint (Hellopeter, Trustpilot, Complaintsboard). A healthy mix has all three. A brittle one is all own-channel or all complaint.
Why it matters: when AI engines reach for the complaint bucket disproportionately, the answer about you tilts negative regardless of how good your product actually is. The fix is editorial spread, not Hellopeter suppression.
How different is your answer in ChatGPT vs Gemini?
Definition: the gap between your best-performing engine and your worst. A small spread means you have a consistent AI footprint; a large spread means you have a structural weakness in at least one engine. For most SA brands the gap is wider than they expect — and it almost always favours either the heritage editorial engine (ChatGPT) or the consumer-forum engine (Gemini), rarely both.
Why it matters: there is no such thing as “the AI answer” — there are four answers, and your customer might be reading any of them.
Which SA sites are shaping the answer about you?
Definition: the ranked list of SA domains currently cited inside answers about your brand, classified by role — default authority, freshness layer, complaint signal, long-tail specialist. This is the operational map for AEO work because it tells you where to publish, where to pitch and where to invest in editorial relationships.
Why it matters: the SA citation core changes per category. Banking lives on BusinessTech and Moneyweb; telecom lives on MyBroadband; e-commerce lives on Hellopeter. Without the source map you are guessing.
Those five metrics are the AEO scorecard. We compute them per brand in the Scorecard and per industry in the Q2 Index Report. The measurement design — blind probes, named-prompt control, Latin-square order-reversal, replication — lives at /methodology and is pre-registered on OSF.
Why SA AEO isn’t US AEO.
The biggest mistake we see is SA brands buying tools built for the US market and assuming they translate. They don’t.
The SA citation core looks nothing like the US one. When AI engines answer questions about SA brands, they reach disproportionately for a small set of South African sites: BusinessTech for commercial commentary, Hellopeter for consumer signal, MyBroadband for telecom and tech, Moneyweb for finance, TechCentral for strategic tech. That is the SA core. Around it sit a rotating cast of category specialists — Cars.co.za and AutoTrader for automotive, RateWeb for banking comparisons, Property24 for real estate.
The amount of citation share each industry routes through SA sources varies dramatically. Short-term insurance is close to entirely SA-cited; AI engines treat it as a hyper-local category and rarely reach for international sources. Safari lodges sit at the other extreme — barely any SA citation share — because the SA specialist publisher set for that vertical is thin and AI defaults to international travel platforms instead. Industry SA-source share runs from over 70% (short-term insurance) to about 3% (safari lodges). Your SA-share number tells you how steeply the AEO playbook tilts toward defending a SA citation core versus pushing into international platforms.
Then there is the Hellopeter asymmetry, which is the single most under-discussed pattern in SA AEO. Gemini reads Hellopeter heavily and weights the complaint signal into its answers. ChatGPT and Claude treat Hellopeter as a low-trust source and effectively ignore it. The same goes for Reddit — Gemini cites Reddit hundreds of times on SA queries; ChatGPT and Claude cite it zero times. So a brand with active complaint volume on Hellopeter or Reddit can look healthy in ChatGPT and bleed in Gemini, which is the engine that powers Google AI Overviews and reaches the largest SA consumer audience.
Multilingual signal also matters. The engines reason about SA brands across Afrikaans, isiZulu, Xhosa and English in ways no US-trained tool measures. An Afrikaans-language review on Netwerk24 or a Zulu-language coverage piece on a regional outlet carries real weight in answers to English-language SA prompts. A measurement framework that only reads English citations is missing roughly a third of the picture.
And the SA regulatory and trust context is its own layer. The Council for Medical Schemes, the FSCA, the Prudential Authority, ICASA — AI engines reach for these as authority anchors when SA category questions get technical. A medical aid brand that shows up cleanly on CMS coverage gets a trust-anchor lift that no global tool would model. A bank that shows up cleanly on PA and SARB coverage gets the same. This SA institutional layer is the cheapest AEO move most enterprise brands have not made yet.
The honest version: SA AEO needs SA measurement, SA editorial relationships and a SA-trained taxonomy. That is what the SA-AEO-Bench is for — a SA-specific benchmark, run quarterly, against the 100 brands and 12 industries SA enterprise actually competes in. Per-industry breakdowns live at /industries/banking, /industries/telecom, /industries/medical-aid, /industries/short-term-insurance and onwards through all twelve.
The tools that play this market.
The AEO tools market is loud and mostly American. A short, honest map.
Cited Brands is the SA-specific platform. We measure 100 SA brands across 12 industries against ChatGPT, Claude, Gemini and Perplexity on a quarterly cycle. The methodology is pre-registered on OSF. The bench includes the SA citation core, the Hellopeter asymmetry, the multilingual citation layer and the SA regulator trust-anchor set. Positioned for SA enterprise that needs local measurement rather than US-extrapolated estimates.
International tools the SA market evaluates include Profound, Athena HQ and parts of the Brandwatch suite. Profound is the largest of the AI-visibility specialists and the one most often quoted in SA pitches. Athena HQ operates in the same space at a smaller scale. Brandwatch comes at AEO from the social-listening direction. None of these were built for the SA market and none of them model the SA citation core, the Gemini-Hellopeter axis or the multilingual layer natively — they can be made to, but the work to localise them is non-trivial.
Methodologically the platforms differ in three places. Probe design — blind vs named-prompt, single-shot vs replicated, ordered vs Latin-square-counterbalanced. Citation classification — surface-level vs taxonomy-coded, English-only vs multilingual. Cadence — continuous vs quarterly vs ad-hoc. The Cited Brands choice is replicated, Latin-square-controlled, taxonomy-coded across multiple SA languages, on a quarterly bench cycle with continuous platform monitoring on top. The full methodology lives at /methodology.
For a tool-by-tool comparison aimed at SA buyers, the best AEO platforms for South Africa post is the long read, and the head-to-head pages at /vs/profound and /vs/dataeq run the specifics. Pick the tool that measures the market you actually operate in.
The action ladder.
Five phases. Each one builds on the last. Most SA brands we work with land somewhere in phase one or two and the whole job is to get them up the ladder without skipping a rung.
Run the free Scorecard.
Before any tactical work, get the baseline. The Scorecard names your mention rate per engine, your closest competitor benchmark, the SA citation sources currently shaping the answer about you, and the three moves that matter most for you. Free, under 24 hours.
Get your free Scorecard →Map your SA-source footprint.
With the baseline in hand, audit your editorial and PR footprint against the SA citation core for your category. The Q2 Index Report names every SA domain that carries citation weight per industry, the role each one plays, and the cheap-win long-tail nobody is fighting for yet.
Read the Q2 Index Report →Quarter-on-quarter, on the same bench.
AEO without a baseline drifts. Set the cadence early: re-read the Index Report every quarter, watch your mention rate and citation share trend, treat the SA citation core as a tracked surface, not a one-off PR list. The bench is re-run quarterly so the trend is against a consistent measurement.
See the research →Run a brand-specific walkthrough.
The Scorecard tells you what is wrong. The walkthrough is where you decide what to do about it. A 30-minute session against your actual data: the per-engine gap, the SA citation core for your category, the editorial moves and entity work that will move the next reading. Customer-led, not pitch-led.
Book a 30-minute walkthrough →Continuous platform tracking.
Once the work is in motion, the platform watches it weekly. The signed-in War Room runs the loop — Measure, Investigate, Optimize, Monitor, Report — on your brand against the SA peer set. This is where AEO becomes a managed discipline rather than a quarterly project.
Inside the Platform →Read the per-industry version.
The 12 SA industries we measure each have their own AEO profile — different citation core, different engine spread, different SA-source share, different posture split.
- Banking →
- Telecom →
- Medical aid →
- Short-term insurance →
- Automotive →
- Retail →
- E-commerce →
- Streaming →
- Real estate →
- Restaurants →
- Safari lodges →
- Higher education →
See how named SA brands sit.
Worked examples from the bench. Each brand page is a concrete read of how the AI engines currently answer about that company.
Common AEO questions, answered straight.
What is AEO South Africa?+
AEO — Answer Engine Optimization — is the discipline of showing up in the short list of brands an AI engine names when a South African asks ChatGPT, Claude, Gemini or Perplexity a question in your category. AEO South Africa is the SA-specific version: the SA sites that shape those answers, the SA brands that compete inside them, and the SA language and regulatory context the engines have to reason about. It is its own field of work, with its own measurement and its own playbook.
Is AEO the same as SEO?+
No. SEO is about ranking ten blue links. AEO is about being named inside a three-name answer. SEO measures position and click-through. AEO measures mention, citation share and narrative control. The disciplines overlap on the supply side — both need a domain to be authoritative and well-cited — but they diverge sharply on what counts as a win. You can rank first on Google and still be invisible to ChatGPT. You can be cited by Claude on every prompt and not own a single page-one keyword.
What is the difference between AEO, GEO and LLM SEO?+
They name the same discipline from different angles. AEO — Answer Engine Optimization — comes at it from the user, who is asking a question and getting an answer. GEO — Generative Engine Optimization — comes at it from the engine, which is generating that answer from a corpus. LLM SEO is the working term inside agencies and treats the LLM as the new SERP. In South Africa we use AEO because it foregrounds the user behaviour change — South Africans are asking AI for advice on medical aid, on banks, on cars — and the optimisation work flows from that.
Do I need an AEO agency, or a platform, or both?+
Most SA brands need a measurement platform first and an agency second. Without measurement, agency work is unaccountable. The Cited Brands free Scorecard gives you the diagnostic in 24 hours. The platform tracks it quarter-on-quarter. An agency layer makes sense once the picture is clear and there is a sustained editorial or PR programme to run against the SA citation core. Buying agency work without measurement is the most common mistake we see.
How long does AEO take to show results?+
Faster than SEO, slower than paid search. The frontier engines refresh their answer sets continuously, but the underlying corpus they reason against — BusinessTech, Hellopeter, MyBroadband, Moneyweb — only updates as fast as the SA media cycle. A new editorial placement in BusinessTech moves Claude and ChatGPT within a few weeks. A new Hellopeter pattern moves Gemini within days. A wholesale repositioning takes a quarter or two of consistent, on-strategy citations to settle.
How is AEO measured?+
Five things matter and the rest is noise. Mention rate: does AI name you at all on an organic, unprompted question? Citation share: when AI does cite, what percentage of those citations point at sources friendly to you? Narrative control: are the citations your own pages, editorial coverage, or complaint platforms? Per-engine spread: how different is your answer in ChatGPT versus Gemini? Source map: which SA domains are currently shaping the AI answer about you? The full per-brand version of those five is the Scorecard. The cross-category version is the Index Report.
What does AEO cost in South Africa?+
The free Scorecard is free. The Q2 Index Report is free to download. The Cited Brands platform — the signed-in War Room with weekly tracking, the per-engine diagnostic, the prioritised actions — is enterprise-priced and quoted per brand on the demo call. Agency-led editorial and PR programmes priced separately depending on the SA citation core you are targeting. Most SA enterprise brands run a measurement subscription plus a quarterly editorial cycle.
Which AI engine matters most for SA AEO?+
Gemini. Not because it is the smartest engine — it is not — but because it powers Google AI Overviews, which is what most South Africans actually see when they search. ChatGPT matters for high-consideration B2B and personal-finance decisions. Claude matters for editorial and methodology-driven questions. Perplexity matters for freshness. But Gemini sits inside Google search and that is where SA consumer behaviour still concentrates. A brand that wins ChatGPT and loses Gemini has won the wrong room.
Is AEO a fad?+
No, and the SA market is moving faster than the global one in some categories. Short-term insurance is already an answer-engine-first category in South Africa — when a Discovery or OUTsurance customer asks "best car insurance South Africa" they are asking an AI, not opening a comparison site. Banking is close behind. The behaviour change is structural and the brands that treat AEO as a fad are the brands that will be missing from those answers in 2027.
My brand already does SEO. Do I need to start AEO from scratch?+
No. Most strong SEO foundations carry into AEO — clean structured data, authoritative domains, deep internal linking, fresh content. What changes is the target. SEO targets keywords; AEO targets questions. SEO targets clicks; AEO targets being named. Your existing SEO work is the floor. The AEO programme sits on top of it and changes what you measure, what you publish, and which SA sites you prioritise relationships with.
Why does AEO South Africa need its own discipline?+
Because the SA citation core is structurally different from the US or UK one. BusinessTech, Hellopeter, MyBroadband, Moneyweb and TechCentral carry weight inside AI answers about SA brands that no global tool models. The Gemini-Reddit asymmetry on SA queries — Gemini cites Reddit aggressively, ChatGPT and Claude do not — is uniquely sharp in this market. Multilingual signal (Afrikaans, isiZulu, Xhosa) only shows up in SA-trained measurement. A US tool applied to a SA brand will miss most of what matters.
How does Cited Brands measure AEO?+
We run thousands of structured prompts per quarter across ChatGPT, Claude, Gemini and Perplexity. Five replications per prompt. Latin-square order-reversal to control for position bias. One hundred SA brands across twelve industries. Citations classified against a published SA source taxonomy. Methodology pre-registered on OSF at osf.io/w4az2. The dataset is called the SA-AEO-Bench and powers everything else — the Scorecard, the Index Report, the platform.
What is the Cited Brands Scorecard?+
A free, one-page diagnostic of your brand against the SA-AEO-Bench. You give us a brand name and an email; we run the brand through the bench prompts and return your mention rate by engine, your top SA citation sources, your closest competitor benchmark and the three moves that matter most for you. Delivered in under twenty-four hours. It is the entry point into the discipline.
What is the Q2 Index Report?+
A quarterly cross-industry reading of the SA-AEO-Bench. Per-engine SA-source share, per-industry citation core, the brands moving fastest, the brands quietly losing ground. Free PDF, ungated topline, gated per-industry tables. It is the field reference. The brand-specific cut is the Scorecard; the category-wide cut is the Index Report.
Keep going.
AI search South Africa
Sibling pillar — what AI search actually is and how it is reshaping consumer behaviour in SA.
AI visibility South Africa
Sibling pillar — how SA brands diagnose where they are visible and where they are not.
Q2 SA AI Visibility Index
The cross-industry quarterly reading of the SA-AEO-Bench.
Where SA brands need to get cited
The cross-industry SA-domain priority list for AI visibility.
Best AEO platforms for South Africa
Tool-by-tool comparison aimed at SA enterprise buyers.
Methodology
The SA-AEO-Bench measurement design. Pre-registered on OSF.
Knowledge base
Long-form reference for analysts, CMOs and PR leads in the SA AEO market.
Glossary
Every AEO and AI-search term in one page. Defined for the SA market.
AEO South Africa is the category. You are the question.
The free Scorecard runs the same SA-AEO-Bench measurement against your specific brand. You get the per-engine gap, the SA sites currently shaping your AI answer, and the three moves that matter most for you.