Our project KBI — business news and analysis on Kazakhstan, a new edition every weekday. Read kbi.news →

Insights

Research worth reading

Short summaries of 2026 publications by international research organisations and data on Kazakhstan. Each summary names its source and links to the original.

Kazakhstan in numbers

Kazakhstan · ConsumersOctober 2026

NIQ: Kazakhstanis spend more without buying more

NIQ's consumer optimism index for Kazakhstan reached 112 in Q3 2026 (109 a quarter earlier, 107 a year earlier). FMCG sales at urban retailers rose 13% over twelve months while volumes stayed roughly flat: inflation drove the growth. 80% of shoppers are cutting everyday spending; nearly half prefer to buy shampoo and household cleaning products online.

Source: NielsenIQ, as reported by Bizmedia.kz · Read the original →
Kazakhstan · Business surveysNovember 2026

Freedom Holding Corp. and S&P Global: services PMI for October

Survey of about 250 services companies, October 12–27. The index fell to 50.6 (51.4 in September): the weakest growth in seven months, new orders stopped growing, while job losses accelerated. Expectations for the year ahead are the lowest since December.

Source: Freedom Holding Corp., S&P Global, KASE · Read the original →
Kazakhstan · E-commerceSeptember 2026

Mastercard: Kazakhstan's retail e-commerce market

Online survey of 1,000 consumers and 200 sellers plus 12 focus groups. The market reached ≈KZT 4.9 trillion in 2025 (+28% a year since 2023); marketplaces hold ≈86% of the domestic segment; 67% of shoppers bought from foreign sellers; forecast ≈KZT 11 trillion by 2030.

Source: Mastercard · Read the original →
Kazakhstan · DigitalNovember 2025

Digital 2026: Kazakhstan — internet and social media in numbers

19.5 million internet users, 93.4% of the population; 16.9 million social media identities (80.8%, +12.4% in a year); 27.3 million mobile connections. Ad reach: TikTok 16.9 million, Instagram 13.1 million, Facebook 2.5 million. Baseline figures for planning online samples in the country.

Source: DataReportal (Kepios) · Read the original →

Published before the review period; we consider it still relevant.

Global practice

Consumer researchOctober 2026

Ipsos Global Consumer Confidence Index, October 2026

Online survey of 21,000+ adults in 30 countries, September 25 – October 9. The index fell 1.1 points to 47.3, below its level a year ago, with the sharpest declines in the U.S. and Europe. Fourteen countries show a significant drop over the year; only six show an increase.

Source: Ipsos · Read the original →
Research industryOctober 2026

Voxpopme and MRII: the state of insights

Survey of 126 research and insights leaders. AI is nearly universal — only 1% use it for none of the listed tasks — yet 51% would not trust AI with any research task without human review. 28% discarded or re-fielded a study over data quality in the past year; fraudulent respondents and AI-generated answers top the industry's concerns.

Source: Voxpopme, MRII · Read the original →
AI in researchAugust 2026

HBR: the innovation problems AI can't solve

Researchers from Harvard, Wharton, Kellogg and Columbia: simulated (synthetic) customers speed up testing but miss the irrational behaviour that shapes real adoption. Advice: keep direct customer contact where it matters most.

Source: Harvard Business Review · Read the original →
Industry trendsJune 2026

GreenBook GRIT 2026 Insights Practice Report

The annual industry benchmark (fieldwork Q1 2026). 70–88% of researchers regularly use fraud-detection tools; the share of brands operating their own panels rose by 10 percentage points year on year; synthetic data moved into the industry's top three topics.

Source: GreenBook · Read the original →
AI and the consumerJune 2026

HBR: how people are really using AI in 2026

Third wave of the AI in the Wild project. Generative AI use keeps broadening, with gradual year-on-year shifts; the authors warn about cognitive offloading and emotional reliance on AI. ChatGPT: 900 million regular users.

Source: Harvard Business Review · Read the original →
Data qualityApril 2026

NORC: fraudulent respondents and bots in online panels

Literature review: industry fraud is estimated at 15–30%, up to 45% on some platforms. An LLM-based synthetic respondent passed 99.8% of standard quality checks; 34% of survey takers admit using AI for open-ended answers. Recommendation: layered defences and verified panels.

Source: NORC at the University of Chicago · Read the original →

Published before the review period; we consider it still relevant.