How South African banks can take advantage of AI search
Mix of local brand sites and international finance comparison sites.
SA-AEO-Bench v1.2 — pre-registered on OSF before data collection. 19,020 successful AI responses across GPT-5, Claude Sonnet 4.5 and Gemini 2.5 Pro. 232,254 cited URLs classified against a published source taxonomy. 100 South African brands, 12 industries. Re-run quarterly. Methodology and analysis code public at osf.io/w4az2.
This page is one industry’s view of that dataset. The per-brand cut is in the Scorecard, the full per-industry tables in the Q2 Index Report.
The pattern.
When you ask AI about SA banking, the answer depends on which AI. ChatGPT defaults conservatively to the Big 5 — Standard Bank, FNB, Absa, Nedbank, Capitec. Gemini elevates TymeBank and Discovery Bank far more readily, especially on "best digital bank" questions. Claude sits in the middle, weighing editorial coverage from Moneyweb and Daily Investor. Perplexity privileges whatever ran in the last fortnight. Four engines, four answer sets, with two or three overlapping names in the middle.
BusinessTech is the SA finance site AI engines reach for first. Moneyweb carries the policy weight. Daily Investor disproportionately shapes the digital-challenger story. RateWeb owns the "cheapest account" question. That is the citation core for SA banking — four sites, plus Hellopeter for one engine in particular. A bank that shows up on the first four is in the answer. A bank quiet on all four is not.
Hellopeter is the SA banking story you cannot ignore. Gemini reads it heavily. ChatGPT and Claude largely do not. That single asymmetry is the biggest per-engine dynamic we measure in SA banking, and it matters because Gemini powers Google AI Overviews — the answer most SA customers actually see. A bank with quiet ChatGPT visibility can still be losing customers to the answer most South Africans encounter. The full per-engine breakdown is in the SA AI Visibility Index Report. The picture for your bank is in the free Scorecard.
Where AI gets its info.
Three to five SA sites carry most of the weight in banking. The names matter. The counts are reserved for the Index Report.
BusinessTech is the SA finance site AI engines reach for first. On rate decisions, on fee comparisons, on "best account" questions — the engines treat it as the SA reference. A roundup placement on BusinessTech lifts you across all four engines because every other SA finance site cites it as the source.
When the question is about Reserve Bank decisions, regulator action, or where SA banking is going next, AI engines reach for Moneyweb. A bank quoted in Moneyweb commentary anchors more confidently in model-led explanations than a bank quoted anywhere else.
Daily Investor coverage of TymeBank, Discovery Bank and Bank Zero shapes the digital-challenger story more than any other source. A digital bank that goes quiet here is the first to drop out of "best digital bank in SA" answer sets.
The Q2 Index Report names every SA domain we measured for banking — the full source taxonomy classified by role (default authority, freshness layer, complaint signal, long-tail specialist) with per-engine citation share for each one. The long-tail is where the cheap wins live. Download the Q2 Index →
For the source map specific to your brand — which domains currently shape AI’s answer about you — run the free Scorecard.
How the engines differ.
Each AI reaches for banking sources differently. One strategy almost never works across all four. The per-engine picture for your brand is in the Scorecard.
ChatGPT names Standard Bank, FNB, Absa, Nedbank and Capitec by reflex. It hedges on TymeBank, Discovery Bank and Bank Zero unless the question explicitly asks for a digital option. Best engine for "established bank" intent. Discounts Hellopeter almost entirely — for now.
Claude weighs Moneyweb and Daily Investor commentary heavier than the other engines. It is the most likely engine to flag a regulatory issue, a leadership change or a strategic shift in the same sentence it recommends a bank. Best for brand-positioning questions. Weakest for transactional "cheapest account" intent.
Gemini pulls Hellopeter complaints and forum threads that ChatGPT and Claude ignore. It elevates the digital challengers more confidently. The most dangerous engine for a bank with active complaint volume, and the engine that shapes Google AI Overviews — which is where most SA customers actually meet AI search.
Recent fee announcements, product launches, rate cuts — Perplexity catches them first. A bank that fed the SA finance press in the last fortnight reads as the current authority. A bank that has gone quiet decays out of this engine first.
Two ways banking brands play this.
Brands in this category split into two camps. Which camp you are in tells you what AI-visibility risks you actually carry.
The Big 5
Standard Bank, FNB, Absa, Nedbank, Capitec. The banks AI engines treat as the safe answer on most questions. Hard to dislodge — they built their AI visibility through decades of SA finance-press coverage. The risk is reputational events. A single regulator action or a big Hellopeter wave shows up disproportionately in Gemini answers, and the customer-perception shift can land faster than the bank’s normal communications cadence can catch up to it.
The digital challengers
TymeBank, Discovery Bank, Bank Zero. These banks built their AI visibility through a low-fee, app-first, branchless story — and the engines reflected the story back, especially Gemini and Perplexity. The win comes through Daily Investor coverage, MyBroadband reviews and a steady drumbeat of editorial. The risk is silence: a digital challenger that goes quiet for a quarter falls out of AI citation sets first, because the engines have nothing else to anchor the brand to.
What changed this quarter.
Questions to take to your next meeting.
Each one is specific. Each one is answerable. We answer them for your brand in the Scorecard and for the category in the Q2 Index Report.
- 01When AI engines answer "best bank for small business in South Africa", is your name in the answer or in the qualifying paragraph?
- 02Where does your bank sit between ChatGPT (conservative, Big 5) and Gemini (Hellopeter-aware)? The gap is almost never small.
- 03Which Hellopeter thread is currently moving Gemini’s answer about your bank? Have you ever read it?
- 04For your closest digital-challenger competitor: which engine treats them most generously, and what coverage explains the lift?
The Q2 Index Report includes the full diagnostic question set for banking — every meeting-ready question with the answer cross-referenced to the per-engine measurement. Plus the SA brands currently ahead, behind, or improving fastest on each one. Download the Q2 Index →
For the answer specific to your brand — where you sit on each question, on each engine — run the free Scorecard.
What to do about it.
Three categories of move lift banking AI visibility. Which specific tactic is right for your brand depends on where you stand today — that conversation lives in the 30-minute walkthrough.
Audit your placement across the five SA finance sites.
BusinessTech, Moneyweb, Daily Investor, RateWeb and Hellopeter form the citation core. A bank on the first four and quiet on the fifth reads very differently to AI than one with active complaint volume. The Scorecard names where your bank stands on each, and which placements have decayed since last quarter.
Quantify the gap between your best engine and your worst before you spend a rand.
A single AI-visibility play for SA banking almost never works across the four engines. The gap is huge — and it is widening. The Index Report quantifies the gap across the sector. The Scorecard quantifies it for your specific bank.
Re-check what AI thinks your bank actually is.
AI engines confuse Standard Bank with Standard Chartered, Absa with Absa Capital Markets, FNB SA with FNB Australia — far more often than the sector realises. Wrong entity, wrong answer. The brand-specific confusions for your bank are in the Scorecard, with the impact of each one.
Brands measured in banking.
Brand-level scorecards for each of these brands. Per-brand visibility figures are in the Index Report; the brand-specific scorecard is one click away.
Common questions,
answered straight.
Why do the Big 5 dominate ChatGPT but lose ground in Gemini?+
ChatGPT learned from a body of SA finance journalism that anchors heavily on the largest banks. It defaults to the brands with the longest cited history. Gemini weighs freshness and forum signal far more heavily, which favours the newer entrants generating disproportionate Daily Investor and Reddit coverage. The full per-engine breakdown is in the Q2 SA AI Visibility Index Report.
How much does Hellopeter actually move AI answers about my bank?+
In Gemini, materially. In ChatGPT and Claude, almost not at all. That gap is the single biggest per-engine asymmetry we measure in SA banking, and it matters because Gemini powers Google AI Overviews — which is what most SA customers actually see. The Hellopeter sensitivity for your specific bank, including which threads currently shape the answer, is in the Scorecard.
Is "best bank for small business" the same answer across all four engines?+
No — it is usually four different answer sets with two or three overlapping names. The banks that appear in all four are the ones actually being evaluated. The banks named by only one are losing customers in the other three. The Q2 Index Report includes the cross-engine overlap for the standard SA banking question set.
How fresh does my bank’s press coverage need to be?+
Perplexity decays fastest — anything older than roughly two months drops out of its current-authority answers. Gemini decays next. ChatGPT and Claude hold for many months. The full per-engine decay-half-life is in the Q2 Index Report. The freshness state for your bank is in the Scorecard.
Where do you get this data?+
We run roughly four to six thousand structured prompts per quarter across ChatGPT, Claude, Gemini and Perplexity. Five replications per prompt. Latin-square order-reversal to control for position bias. Methodology pre-registered on OSF. See the full methodology for details.
Keep going.
Best banks in South Africa, AI-ranked
How the engines actually rank SA banks — and where they disagree most.
How SA banks can improve AI search visibility
The categories of move that lift banking visibility — without giving away the playbook.
Defending AI-search budget to the CFO
A finance-team frame for why AI visibility now matters to revenue.
Banking is the category.
You are the question.
The free Scorecard runs the same banking 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.
- Dataset: sa-aeo-bench-v1
- Snapshot: 2026-05-19 · 17 days old · budget 270 days
- Pre-registration: osf.io/w4az2
- 14,826 responses · 188,877 citations · 100 brands · 10 industries
- POPIA-compliant · Information Officer registered
- Data residency: SA + EU only
- Methodology: pre-registered before data collection · audit trail per record
- Models measured: OpenAI GPT-5 · Anthropic Claude Sonnet 4.5 · Google Gemini 2.5 Pro