Continuous monitoring of your brand across five AI engines. Custom competitor benchmarking. Monthly remediation playbooks. Quarterly board-ready briefings.
See what AI is saying about your brand in South Africa.
Cited Brands measures and improves how ChatGPT, Claude, Perplexity, Gemini and Google AI Mode describe South African brands. Based on the largest open AI citation study in Africa — 196,834 citations across 100 SA brands.
Methodology pre-registered on OSF · 196,834 citations · 100 brands measured · POPIA-compliant
Banking
Banking
BankingBanking
Banking
Banking
BankingBanking
- TymeBankBanking
BankingInsurance
Insurance
Insurance
Insurance
Insurance
- Discovery InsureInsurance
Insurance
- NakedInsurance
- Old MutualInsurance
- SanlamInsurance
- LibertyInsurance
- Auto & GeneralInsurance
- Discovery HealthMedical aid
Medical aid
Medical aid
Medical aid
Medical aid
Medical aid- GEMSMedical aid
Medical aid
- BankmedMedical aid
Medical aidTelecom
Telecom
- Cell CTelecom
Telecom
Telecom
Telecom- WebafricaTelecom
Telecom- RSAWebTelecom
Telecom
- ShopriteRetail
- CheckersRetail
Retail
RetailRetail
Retail
- GameRetail
Retail
Retail
RetailE-commerce
- SuperbalistE-commerce
E-commerce
- ZandoE-commerce
- Bid or BuyE-commerce
- LootE-commerce
- Faithful to NatureE-commerce
Automotive
Automotive
Automotive
AutomotiveAutomotive
Automotive
Automotive
Automotive
Automotive
- Cars.co.zaAutomotive
Streaming
Streaming
Streaming
Media
MediaReal estate
- SeeffReal estate
Real estate
- Property24Real estate
Real estate- Rawson PropertiesReal estate
- Chas EverittReal estate
Restaurants
Restaurants- SpurRestaurants
- SteersRestaurants
- Mugg & BeanRestaurants
Restaurants
Restaurants
Restaurants- Ocean BasketRestaurants
- SasolIndustrial
- NaspersIndustrial
- ProsusIndustrial
IndustrialIndustrial
IndustrialIndustrial
- SABIndustrial
Industrial
Industrial
Banking
Banking
BankingBanking
Banking
Banking
BankingBanking
- TymeBankBanking
BankingInsurance
Insurance
Insurance
Insurance
Insurance
- Discovery InsureInsurance
Insurance
- NakedInsurance
- Old MutualInsurance
- SanlamInsurance
- LibertyInsurance
- Auto & GeneralInsurance
- Discovery HealthMedical aid
Medical aid
Medical aid
Medical aid
Medical aid
Medical aid- GEMSMedical aid
Medical aid
- BankmedMedical aid
Medical aidTelecom
Telecom
- Cell CTelecom
Telecom
Telecom
Telecom- WebafricaTelecom
Telecom- RSAWebTelecom
Telecom
- ShopriteRetail
- CheckersRetail
Retail
RetailRetail
Retail
- GameRetail
Retail
Retail
RetailE-commerce
- SuperbalistE-commerce
E-commerce
- ZandoE-commerce
- Bid or BuyE-commerce
- LootE-commerce
- Faithful to NatureE-commerce
Automotive
Automotive
Automotive
AutomotiveAutomotive
Automotive
Automotive
Automotive
Automotive
- Cars.co.zaAutomotive
Streaming
Streaming
Streaming
Media
MediaReal estate
- SeeffReal estate
Real estate
- Property24Real estate
Real estate- Rawson PropertiesReal estate
- Chas EverittReal estate
Restaurants
Restaurants- SpurRestaurants
- SteersRestaurants
- Mugg & BeanRestaurants
Restaurants
Restaurants
Restaurants- Ocean BasketRestaurants
- SasolIndustrial
- NaspersIndustrial
- ProsusIndustrial
IndustrialIndustrial
IndustrialIndustrial
- SABIndustrial
Industrial
Industrial
Independent research. Open methodology. No brand pays for placement.
AI is the new search layer for South African consumers — and most brands are flying blind in it.
When South Africans search for the best bank, the best medical aid, the best car insurance, the best place to buy a phone, they no longer just open Google. They open ChatGPT, Claude, Gemini or Perplexity. These engines decide which brands get mentioned, which get recommended, which get ignored — and right now, your brand probably has no idea what they’re saying.
of AI citations for SA queries come from South African sources
of comparison answers flip when you reverse the word order
is how often Bidvest Bank, Toyota SA and Pick n Pay appear in unprompted "best X in SA" queries
Every marketing funnel has four stages. AI has the same four. Most measurement tools track only one.
The Cited Brands platform measures all four stages of how AI engines describe your brand. Each broken stage has different causes — and different fixes.
| Stage | What it measures | What “winning” looks like | What “broken” looks like |
|---|---|---|---|
AWARENESS | Does AI recommend you when nobody mentions you? "What’s the best [category] in SA?" | AI lists 5–8 brands and you’re one of them | AI never mentions your name (Bidvest Bank scores 0% here) |
KNOWLEDGE | Does AI know about you when prompted by name? "Tell me about [your brand]" | AI returns accurate, complete description | AI confuses your brand with competitors or returns outdated info |
CONSIDERATION | Does AI rank you against named competitors? "Compare [you] vs [competitor]" | AI picks you on at least one dimension | AI picks the competitor every time |
PURCHASE | Does AI describe your value accurately? "Is [your brand] worth it?" | AI repeats your positioning correctly | AI cites the wrong price, wrong product mix, wrong target customer |
Bidvest Bank’s Knowledge score is 96%. Their Awareness score is 0%. These need different fixes. Our platform tells you which.
We measure. We diagnose. We fix.
One dashboard. Five AI engines. Every week.
Every week we run hundreds of prompts about your brand and your top competitors across ChatGPT, Claude, Gemini, Perplexity and Google AI Mode. We test all four stages of how AI describes you — awareness, knowledge, consideration, purchase — and show you exactly where you sit.
See exactly which sources control your AI narrative.
We map which sources are controlling your AI narrative. Which SA publishers (BusinessTech, MyBroadband, HelloPeter) are mentioning you positively. Which international complaint sites (Trustpilot, Pissed Consumer) are dragging you down. Which competitors AI prefers over you and why.
We move the score for you.
Every month our team executes the highest-leverage actions on your behalf — the specific levers that shift each engine’s behaviour stay proprietary to our customers, because they’re the moat. For enterprise customers, we ship the top three moves each month and report what landed. You watch your score climb quarter over quarter. The playbook is the product.
Built on the largest open AI citation study ever published in Africa.
SA-AEO-Bench v1: 196,834 citations classified across 100 SA brands and 10 industries. Methodology pre-registered on the Open Science Framework before data collection began. Bootstrap 95% confidence intervals on every measurement. Latin Square debiasing on every comparison.
Read the headline findings
The seven hypotheses, what we confirmed, what surprised us.
Download the full 23-page report
Every chart, every confidence interval, every per-industry breakdown.
View the OSF pre-registration
The protocol we registered before any data collection started.
We publish what we measure. We can’t show every operational detail — some prompts and scoring stay proprietary — but we show enough that any researcher can verify our methodology.
We test how AI responds in Afrikaans and isiZulu — because no other AEO tool does.
1,500 multilingual API calls. Same SA category questions, asked in your customer’s first language. The two markets behave fundamentally differently — and most SA brands have no idea.
isiZulu is the most-spoken first language in South Africa. SA brands targeting isiZulu audiences via AI search have zero native-language editorial control over what AI surfaces. Afrikaans-targeted brands have a real channel. We tell you which one you’re in.
From input to action in four steps.
- 01
You tell us your brand and category.
Brand name, domain, top 5 competitors.
- 02
We run our citation protocol weekly across the 5 major AI engines.
ChatGPT, Claude, Gemini, Perplexity, Google AI Mode.
- 03
You get a weekly intelligence brief showing exactly what AI is saying.
Score deltas, source map, competitor moves.
- 04
We tell you what to change — and execute it for you if you’re on Enterprise.
Ranked playbook. Top three actions shipped each month.
Built for the people accountable for how brands show up.
Track every AI response that mentions your brand. Catch competitive threats 90 days early. Defend category position in every engine.
Find out where AI sources its negative answers. Engage the international complaint platforms your local PR currently misses.
Monitor how AI describes your brand to analysts and institutional researchers. Know what AI Overviews returns before your earnings call.
Pricing tailored to your portfolio and AI engine coverage.
Three tiers: Starter, Growth, Enterprise. Mid-market to JSE Top-100 portfolios. POPIA-compliant. 90-day money-back guarantee. We price every plan in Rand and against your specific brand count and competitor set — talk to us for an exact number.
Common questions
How is this different from Brandwatch, Cision or DataEQ?
Those platforms measure social media, news, and reviews. We measure AI engines specifically — ChatGPT, Claude, Gemini, Perplexity. Different data, different methodology, different findings. We complement rather than replace traditional reputation monitoring.
Do you actually run prompts on ChatGPT, Claude and Gemini?
Yes. Every week we run hundreds of standardised prompts across all five major AI engines using their official APIs. We classify every cited URL. We aggregate the results. The methodology is published in full at osf.io/w4az2.
What if my brand isn’t in your research yet?
We’ve measured 100 SA brands in v1. If your brand isn’t on the public index, we can add you with a 4-week onboarding (included in Growth and Enterprise tiers). Starter customers can request brand addition in our Q3 2026 update.
How long does setup take?
1 week. We get your brand details, competitor list and target queries, configure the monitoring, and your first weekly brief lands at the end of week 1.
Is the data POPIA-compliant?
Yes. All brand data is processed in compliance with POPIA. Data residency is in South Africa or EU jurisdictions only. We don’t process personal data — only public AI engine outputs about brands.
Can I export the data?
Yes. CSV and JSON export on every tier. Enterprise customers get full API access.
What happens if AI changes how it works?
We re-run methodology validation every quarter. When a major LLM provider updates their model, we re-baseline. Your reports always reflect the current state of AI.
What’s the difference between the free research and the platform?
The free SA-AEO-Bench is a quarterly snapshot of 100 SA brands — public, methodology pre-registered. The platform is continuous weekly tracking of YOUR brand specifically, with custom competitor sets, remediation playbooks, and execution. Most customers start with the research, then upgrade.
Find out where your brand ranks in AI search.
Free 60-second snapshot. No credit card. No sales call.
Run my SA AI Visibility Snapshot →Powered by SA-AEO-Bench v1 · 196,834 citations · 100 brands · ~30 seconds, no signup.
Your scorecard takes 60 seconds. If you want to talk afterwards, we make it easy. If you don’t, we don’t chase.