A new survey by Ipsos has exposed a troubling pattern across Nigeria's small business sector: most enterprises that experiment with artificial intelligence abandon the technology before it delivers any real value. The findings paint a stark picture for the country's 40 million registered SMEs, which collectively employ roughly 60 percent of Nigeria's working population.
The Scale of the Abandonment
Ipsos spoke with more than 500 small and medium business owners across Lagos, Abuja, and Port Harcourt during the third quarter of this year. The results showed that 87 percent of firms that piloted an AI tool in the past 18 months had stopped using it by the six-month mark. Only 13 percent reported ongoing use, and of those, just a fraction described themselves as satisfied with the outcomes.
Google Nigeria, which has run its own skills development programmes in partnership with local tech hubs, confirmed the trend matches what it hears from entrepreneurs. "The gap between trying and succeeding is still very wide," a spokesperson told reporters in Lagos.
What Goes Wrong in the First Weeks
Business owners cited several recurring problems.setup costs overwhelmed smaller operations that had budgeted for monthly subscription fees alone. In Lagos's Ikeja district, where dozens of digital freelancing startups cluster around the technology park, founders described spending weeks on integration only to see the system produce inaccurate customer responses.
Segun Afolayan, who runs a logistics firm in Owerri with 12 employees, spent roughly 350,000 naira on an AI chatbot service last March. By July, he had cancelled the contract. "It could not read our local addresses properly," Afolayan said. "Customers were giving orders and the system did not understand half of what they typed."
Training Gaps and Language Barriers
Language emerged as a significant challenge beyond the technical setup phase. Most commercial AI tools are optimised for English, yet a large share of customer interactions in Nigeria happen in Pidgin or local languages such as Yoruba, Igbo, and Hausa. When the AI fails to understand or respond appropriately in those tongues, customer frustration rises quickly and owners lose trust in the platform.
Training presents another barrier. Smaller firms rarely have staff with the technical background to configure AI tools correctly or to interpret the data these systems generate. Larger competitors can absorb those costs, but for a company of five people, diverting one worker to manage an AI dashboard creates an immediate productivity problem.
The Cost Problem Hits Hard
Money plays a central role in the retreat from AI. Several SMEs told Ipsos that they signed up for low-cost trial periods, assuming the free tier would be sufficient to test whether AI made sense for their operations. When the trial ended, the jump to paid plans — often priced in dollars — proved unaffordable given the naira's volatility against foreign currencies.
Abuja-based retail consultant Funke Adeyemi described a common scenario. "The tool looked great for two weeks. Then they wanted $49 a month, and with the exchange rate what it is, that is suddenly not a small experiment," she said. "For my kind of business, it became a big expense before I had seen a single extra sale from it."
What Nigerian SMEs Actually Needed
The survey data suggest that many SMEs downloaded or subscribed to AI tools without a clear plan for how the technology would improve their revenue or reduce costs. Nearly half of respondents admitted they did not set specific targets before starting their pilot. That lack of direction made it easy to pull the plug when early results disappointed.
Support structures also remain thin. Google's.ng digital skills initiative has trained more than 100,000 Nigerians since 2019, but the programmes focus primarily on basic digital literacy rather than advanced AI implementation for business owners. Industry observers argue that more needs to be done to bridge the gap between tool availability and practical, affordable integration support.
Regional Inequality Deepens the Problem
The Ipsos findings also reveal a geographic divide. SMEs in Lagos showed considerably higher persistence rates with AI tools compared with firms in the north or in rural areas. Connection speeds, access to reliable electricity, and proximity to technology communities all contribute to the gap. In states like Katsina and Adamawa, entrepreneurs reported that basic internet connectivity remained unreliable, making sophisticated cloud-based AI services practically unusable.
This inequality means AI adoption, rather than levelling the playing field for Nigerian small businesses, risks reinforcing existing economic divides between urban and rural operators.
Who Is Still Succeeding and Why
Not every Nigerian SME is giving up. A subset of firms, most of them in Lagos and a few in Ibadan, reported that targeted use of AI for specific tasks such as inventory management and social media scheduling had produced measurable results within three months. These businesses shared certain traits: they started with a narrow use case, ensured at least one staff member received basic training, and set clear benchmarks for success before committing to paid subscriptions.
Yinka Bello's food distribution company in Lagos is one example. Rather than adopting a comprehensive AI customer service tool, she used a simple AI-powered inventory tracker that integrated with her existing accounting software. "We spent less than 80,000 naira total," Bello said. "It tells us when stock is running low. That alone saved us from two costly supply mistakes last quarter."
What Happens Next
Policymakers at Nigeria's Federal Ministry of Industry, Trade and Investment have signalled interest in addressing the support gap for SME technology adoption. A spokesperson confirmed that the ministry is reviewing proposals for subsidised AI integration programmes, though no timeline for implementation has been announced.
For now, the evidence suggests Nigerian small businesses need more than access to powerful tools. They need affordable, locally relevant implementation support, better training options, and pricing structures that account for the pressures of operating in a volatile currency environment. Until those conditions improve, the rate of abandonment shown in the Ipsos survey is unlikely to shift significantly, and millions of entrepreneurs will continue to quietly quit AI without ever seeing its promised benefits.
