Linkedin
Facebook
Email
more
In 2020, a joint study ("Report") by the Singapore Institute of Technology, RSM Singapore, and the Institute of Singapore Chartered Accountants found that 69% of Singapore SMEs hadn't adopted data analytics. Many were still relying on basic spreadsheets, with little understanding of what more advanced tools could offer them.
That was before the AI boom. Between 2022 and 2024, generative AI tools became cheap, accessible, and able to process information through plain English prompts. With technical barriers seemingly lower than ever, you might expect adoption to have climbed sharply since. It hasn't. A 2025 study by researchers at the Singapore University of Social Sciences cites government data showing that only 12.5% of Singapore SMEs have adopted big data analytics, compared to 82% of large enterprises. The two figures aren't directly comparable, since they come from different studies measuring slightly different things. Still, the direction is hard to miss. Years after AI made analytics more accessible than ever, most SMEs still haven't made the shift.
This raises an obvious question. If the tools are cheaper and easier to use than ever before, why hasn't adoption followed? The answer is that AI has lowered the entry barrier, but it hasn't removed the real obstacles. AI can't function without clean, organised data to work with, and building that foundation remains a heavy operational lift for smaller firms. This has little to do with how good or affordable the AI tools themselves have become.
The problem starts beneath the data
The first obstacle sits underneath the data itself, long before any AI tool gets involved. The Report found that a lack of information systems infrastructure, along with a shortage of in-house data expertise, were among the top reasons SMEs held back from adopting data analytics in the first place. In practice, this often means data is scattered across disconnected spreadsheets, individual employees' inboxes, or whatever system was fastest to set up at the time. There's rarely a single, standardised source of truth.
This matters even more now that AI is in the picture. Feeding an AI model fragmented, inconsistent data doesn't just produce mediocre results. It produces outputs that look polished while being wrong. A chatbot or analytics tool will still generate a clean-looking chart or a definitive-sounding answer, regardless of whether the underlying numbers were ever reliable. An AI tool that visibly fails isn't the real danger. The real danger is one that fails quietly, while still looking right.
This is why adopting AI can't substitute for the harder work of organising a business's data in the first place. Tidying up fragmented records, deciding on consistent formats, and establishing a single reliable source of data is unglamorous work that has little to do with AI at all. But without it, AI has nothing solid to work with, no matter how advanced the model.
The hidden cost of getting data-ready
That same shortage of in-house expertise doesn't disappear once a firm decides to fix its data. It simply resurfaces as a cost. Shifting to a data-driven way of working means restructuring how you collect information, retraining staff on new habits, and rethinking workflows that have often been in place for years. For a small team where everyone is already stretched across daily operations, this isn't a minor task.
This explains why government support hasn't been enough to move the needle. A 2025 SUSS study found that generic government advocacy and subsidy levels weren't significant predictors of whether SMEs actually adopted big data analytics. What mattered more was organisational readiness: having the internal skills, infrastructure, and leadership buy-in already in place. In other words, access to support means little without the internal capability to use it.
The same logic applies to AI. Simply having powerful tools available doesn't help a firm that lacks people who know how to put them to work, whether that expertise is hired in or built up through training existing staff. Building that capability is expensive either way. Even Accenture, a professional services firm with hundreds of thousands of employees and far greater resources than any SME, disclosed $923 million in restructuring charges as part of its push to embrace AI, including reskilling staff and overhauling its operating model. For a lean SME with a tighter budget, this is likely to be a significant barrier.
Where AI genuinely helps
None of this means AI has nothing to offer SMEs. Once you've organised the underlying data, AI becomes a genuinely useful tool. Not a magic fix, but a way to draw more value from work you're already doing well.
The path there doesn't require overhauling every department at once. A more realistic approach starts small:
- Audit how data currently flows through the business.
- Identify where the biggest errors or inconsistencies originate.
- Fix that one point before expanding further.
Cleaning up data at the source, before it ever reaches a spreadsheet or dashboard, matters more than introducing analytics tools further down the line. Only once a single process is reliably clean should you expand the same discipline to other parts of the business.
This also reframes what AI adoption should look like for an SME. Getting there looks less like a single decision to 'get AI' and more like a series of smaller, deliberate steps that build on each other. A firm that does this ends up in a far stronger position than one that buys an expensive analytics platform and points it at everything at once. AI doesn't create order out of chaos. It amplifies wherever discipline already exists.
How RSM Singapore can help
This is exactly where we partner with you. Digital transformation is the strategic integration of digital technologies into all areas of a business to improve operations and deliver value to clients. For many business owners, the journey can feel complex. We help you move from identifying operational pain points to implementing fit-for-purpose digital and analytics solutions that unlock value and future-proof your business.
Explore our digital transformation and advisory solutions - https://www.rsm.global/singapore/service/digital-transformation-services-business-diagnostics-process-re-engineering-advisory
The real competitive advantage
The rise of accessible AI has made data analytics look like something any business can simply switch on. In reality, the barriers that have kept SME adoption low for years haven't gone away. They've just become easier to overlook. Clean, organised data and the people to manage disruption remain what a firm actually has to invest in to compete, regardless of how good the underlying AI has become.
This matters because most firms will eventually have access to AI. The gap between them will come down to something else entirely: who spent the time building a structured data environment capable of putting that access to use. As AI becomes further embedded into how business gets done, the question worth asking is how long this foundation stays a competitive advantage before it simply becomes the bare minimum needed to compete.
