How Kaspi.kz built a profitable super-app in a 20-million-person market

Executive overview

Kaspi.kz started as a struggling Kazakh bank in 2007 and became a $20B Nasdaq-listed super-app processing $80B in payments annually — with 14 million monthly active users in a country of 20 million people. The company competes across payments, e-commerce, fintech, travel, grocery, and government services from a single app under a single brand.

The constraint of a tiny market made product quality non-negotiable: lose a million unhappy users and the growth story ends. Every product lives or dies by its net promoter score, and the culture of extreme simplification — launch one feature, add the next quarterly — compounds into an innovation machine 50 product teams wide.

In a small market, quality isn't a virtue — it's survival.

From bank to super-app: the evolution

• Entered financial services 2007–08; built the leading consumer lending and savings products in Kazakhstan
• Moved sequentially into payments, e-commerce (2014), travel, B2B payments, grocery, government services
• Chose the super-app model early, against industry consensus that single-purpose apps were optimal
• Mission — "improve people's lives by developing world-class mobile services" — deliberately contains no industry or product category, allowing expansion into any vertical
• Every new vertical entered with zero prior experience; treated as a competitive advantage, not a liability

The NPS-driven operating model

• Half a million consumers surveyed monthly on each product
• Callbacks made to dissatisfied users; product teams listen to verbatim recordings
• Each product has a directly accountable owner and team; no company-wide financial KPIs for product people
• A negative NPS triggers a kill decision — not a fix-it programme
• Killed a credit card worth ~$200M in net income contribution because it couldn't reach NPS-positive; described internally as "bad money"
• NPS on e-commerce services reaches 90; financial products run 50–60

Payments: the closed-loop network

• P2P payments (similar to Venmo/Cash App): instant, commission-free transfers that drove Kazakhstan's shift from 85–90% cash in 2016 to 90% cashless today
• QR payment network: merchant-facing, priced below Visa/Mastercard rates; Kaspi controls both sides of the transaction
• B2B payments: invoice settlement for convenience stores and merchants receiving supplier deliveries
• Closed-loop design means money moves within Kaspi's system, cutting out card network intermediaries

Marketplace: the Amazon of Kazakhstan

• Launched 2014 in electronics; now 8M+ monthly actives, 8M+ SKUs across all categories
• Structured around the product, not the seller: multiple merchant offers per SKU, enabling price comparison
• Kaspi does not compete with retailers — it gives offline merchants logistics, advertising, working capital financing, and a nationwide sales channel
• Delivery is largely free; most orders fulfilled same-day or next-day
• E-grocery launched ~2023: 10,000 data-selected SKUs, three-hour delivery window, profitable within 12 months at 6–7% net income margin while still growing ~100% YoY
• Grocery economics work because average basket is ~$30 (weekly shop), not quick-commerce impulse buys

Fintech: lending that fuels the flywheel

• Buy-now-pay-later decisions made in under one second; 99.9% automated
• Cost of risk below 2% — described as world-class for consumer lending
• BNPL increases consumer wallet size → drives marketplace GMV
• Working capital loans to merchants → merchants buy more inventory → GMV grows further
• No credit cards; the product was killed when it failed the NPS test

Government services and digital infrastructure

• Kazakhstan moved from near-zero digital government to issuing driving licences, registering businesses, filing taxes, transferring car ownership — all in-app
• Kaspi holds no exclusivity; government services can run on other apps, but Kaspi's UX means most citizens default to it
• Driven by a political digital-reform agenda under President Tokayev and Kaspi's willingness to partner with agencies directly

Extreme simplification as a management system

• New products launch with the single most important feature only; additional features added one per quarter
• Benefits: smaller launch team, simpler consumer onboarding, fewer support queries, lower costs, faster path to profitability
• 50 product teams each shipping at least one feature per quarter = ~50 new features every quarter across the platform
• Profitability is a byproduct of simplicity, not a separate goal

Growth outlook and international expansion

• Kazakhstan opportunity: services retail (barbershops, plumbers, electricians) and physical retail verticals still underpenetrated; both sectors forecast to grow ~14% annually for five years
• Kaspi targets growing faster than the market by going vertical-by-vertical with existing 14M users and 7,000+ merchants as the distribution base
• Marketplace revenue guided to grow 65% in current year
• Acquired 65% of HepsiBurada, one of Turkey's largest e-commerce platforms (85M-person market); plan to introduce Kaspi's fintech and payments stack into Turkey
• Listed on Nasdaq January 2024, delisted from London Stock Exchange; stock liquidity increased ~20x post-move

Why super-apps fail elsewhere

• Large markets mask poor product quality with user-acquisition growth
• When growth stalls, negative network effects surface: unhappy users churn, re-acquisition costs 10x more than retention
• Kaspi had no such buffer: in a 20M-person market, 1M dissatisfied users is existential
• Foundational products that consumers love are the prerequisite; network effects only compound quality, they do not substitute for it