We Analysed 25 African Banks' Digital Strategies. Most Don't Disclose the Numbers That Matter.

How We Did This

We selected 25 banks covering Africa's largest retail banking markets — South Africa, Kenya, Nigeria, Egypt, Tanzania, Morocco and the pan-African groups — and, in October 2026, rechecked each against its latest primary-source document: annual report, results presentation, or exchange filing (JSE SENS, NGX, AMMC). Periods range from H1 2025 to FY2026 because year-ends differ.

Each bank was rated on four questions: (1) does it quantify an external distribution network — agents, telcos or merchants? (2) does it report digitally active clients separately from total clients? (3) does it quantify digital lending — disbursements or book? (4) does it separately report non-bank ecosystem income — insurance, telco, asset management, lifestyle?

Verified means the document states a figure. Partial means it describes the activity or gives a proxy. Unconfirmed means we did not find it in the document — *not* that the bank does not do it. Equity, KCB and M-Pesa partnerships, for example, are widely known; what we are measuring is what banks put in their results.

Limitations: some long reports were reviewed through their results sections and presentation slides rather than every page; press releases outside the results documents are noted but do not count as Verified. We will update ratings if a bank points us to a figure we missed.

1. The Disclosure Gap Is the Headline

We set out to count which banks do what. The more important finding is what they *report*. Across 100 bank-by-metric checks, only 8 are Verified in primary results. Most banks headline deposits, loans and profit — and describe digital in narrative without the figure.

That matters commercially. If the board pack shows registrations and transaction volumes, the organisation optimises for activity. Six of the banks we checked — KCB, Zenith, FirstBank, CIB, NBE and Absa Kenya among the least specific — give investors no digital metric we could rate Verified at all.

2. Engagement Is Not Activation

Where banks do report digitally active clients, the numbers show how far registrations overstate reach.

Capitec's audited FY2026 results report 15.3m active app clients against 25.8m active clients — roughly 59%. Absa Group reports 5.4m digitally active against 13.1m active customers (about 41%). TymeBank's Pillar III disclosure shows 3.7m active customers out of a base of about 11.3m — a reminder that even a digital-only bank has a large inactive tail.

Others give figures that look similar but aren't: UBA reports 44.8m "digital banking customers" without saying how many are active; Ecobank reports 12.8m active consumers without isolating digital; Attijariwafa reports 444m digital connections and a 94.4% digital share of transactions — usage, not clients.

3. Digital Credit: One Bank Shows the Numbers

Transaction fees are under pressure from instant payment schemes — PayShap in South Africa, NIP in Nigeria, InstaPay in Egypt. The banks that can build durable digital revenue are those turning transaction data into credit decisions.

Only NCBA quantifies this in its results: KES 646bn in digital loans disbursed in H1 2025, up 35%, and more than KES 1tn disbursed across Africa — largely through M-Shwari and Fuliza with Safaricom. That is distribution and credit working together: the telco supplies reach, the bank supplies the balance sheet and the decisioning.

Capitec (AvaFin online lending) and Access Holdings (Oxygen X, ₦19bn+ loan value) disclose digital lending subsidiaries without a group-wide digital book. Every other bank reports only total loans. Given the Central Bank of Kenya's digital credit provider rules and tightening elsewhere, investors should expect — and ask for — more.

4. Distribution: Visible in Strategy, Rarely Sized in Results

Distribution partnerships run through many of these banks' strategies, but few size them in results. Co-operative Bank of Kenya is the clear exception: 16,055 Co-op Kwa Jirani agents and 619 Sacco front offices.

Others describe the model without the number: FNB's AgencyPlus agent stores and FNB Connect retail distribution; TymeBank's retail kiosks; African Bank's "Alliance Banking" partnerships; UBA's 430,000 POS terminals; NCBA's Safaricom-linked lending. We rate these Partial. Own branches and ATMs — however large, like Attijariwafa's 5,593 Moroccan branches — are not counted as external distribution.

5. Ecosystem Income: Concentrated and Lightly Reported

Three banks separately quantify income beyond core banking: Capitec (value-added services and Capitec Connect, R6.1bn, +38%), NCBA (non-bank subsidiaries 15% of group profit before tax) and Co-operative Bank (funds under management KES 461.7bn, plus insurance, brokerage and leasing lines).

Many more mention insurance, asset management or payments verticals — Discovery, GTCO, Stanbic IBTC, Ecobank, Absa Kenya — without breaking out the income. Ecosystem revenue is attractive because it is fee income not tied to the rate cycle, but it can only be judged if it is reported.

6. Architecture and Launch Speed: Not Proven

We set out to test whether banks on unified platforms with an API layer launch products faster. Several describe cloud and API programmes, but none publishes comparable time-to-market data, and we found no independent study covering these banks. The claim is plausible; we cannot evidence it, so we do not make it.

The Takeaway

Stop asking "how many users are on our app?" Start asking: what does a digitally activated client actually earn us — and through which distribution partner did we reach them?

For bank executives that means three changes: report digitally active clients, not registrations; size agent, telco and merchant channels in the results, not just the strategy slides; and disclose digital credit volumes and losses. The banks that already do — Capitec, Absa, NCBA, Co-op — are the ones investors can actually value on digital.