Beyond Digital Adoption: Why Malaysian Micro and Small Enterprises Need Operational Capability

Athirah Nor Kamarudin

Abstract

Malaysia’s micro and small enterprises are increasingly connected to digital systems. They accept digital payments, communicate with customers through online platforms, use accounting or point-of-sale applications, participate in e-commerce, and increasingly encounter regulatory systems that require digital interaction. Yet adoption does not necessarily amount to capability. A business may possess several digital tools while continuing to rely on fragmented records, informal workarounds, tacit knowledge and the owner’s personal memory to understand what is actually happening inside the organisation.

Recent research by the Khazanah Research Institute describes Malaysian micro and small enterprise digitalisation as broad but shallow, with adoption concentrated in customer-facing functions and weaker integration into back-end operations. This essay approaches that gap ethnographically. Drawing on my experience working with Malaysian SMEs, manufacturing operations, associations, organisational transformation projects and adult digital training, I ask a slightly different question from the conventional technology-adoption discussion: what must an organisation know, understand and reorganise before a digital tool becomes genuinely useful?

I argue that the central policy problem is not simply insufficient technology adoption. It is the uneven distribution of what I call operational capability: the ability of an organisation to understand what a digital system is asking of it, connect that system to existing working practices, maintain reliable information, interpret what the resulting data means, preserve organisational knowledge and adapt when requirements change. Seen this way, digitalisation becomes not only a technological problem but also an organisational, educational and institutional one.

A screen in Miri

In a training room in Miri, I watched a woman in her fifties sit in front of a laptop and become unusually still. She ran a business and was neither inexperienced nor incapable. Like many small-business owners I have encountered, she already understood the practical economics of her livelihood: which customers were dependable, when stock had to be replenished, which supplier could accommodate a last-minute request, when cash was tight and which expenses could be postponed without disrupting the business.

Yet the logic appearing on the screen in front of her seemed to belong to another world.

We were conducting a programme intended to prepare Malaysian SMEs for e-invoicing. The new process required participants to think about familiar commercial activity through a formal digital structure: customer information belonged in defined fields, transactions had to be classified consistently, records needed to follow prescribed formats and the information entered at one point would travel elsewhere through a wider institutional system.

For someone already accustomed to accounting software or structured digital records, such requirements could appear administrative rather than conceptual. For several participants in the room, however, the shift was more fundamental. Their businesses were not devoid of information. On the contrary, information was everywhere. It existed in handwritten records, phone messages, receipts, bank transactions, remembered conversations and routines repeated often enough that they no longer needed to be written down.

What was missing was not necessarily knowledge. What was missing was a translation between forms of knowledge.

My initial assumption was that the challenge was instructional. Explain the interface clearly enough, demonstrate the correct sequence and allow sufficient time for practice, and participants would eventually acquire the necessary skill. The longer I watched people work through the exercise, the less convincing that explanation became.

The difficulty was not simply knowing where to click. Participants needed to understand why the system required particular information, what a classification represented, who would eventually use the data and what would happen downstream if something were entered incorrectly. The system treated these categories as self-evident. To the user encountering them for the first time, they were abstractions imposed upon a business that had previously been understood through experience.

I began asking participants to show me how they already worked. Instead of beginning from the interface, we began from the business. How did they recognise a completed sale? Where did they record it? Which transactions mattered enough to be written down immediately? Which details could be remembered? What happened to a receipt after a purchase? Who knew whether payment had been collected? Where would they look if something did not balance?

Once we started there, the digital exercise changed character. A field on the screen could be connected to an action they already recognised. A transaction category could be explained through something that happened inside their own business rather than through the vocabulary of the software. The sequence stopped being simply procedural and began to acquire an institutional logic.

By the end of the exercise, the most important change to me was not that a participant had successfully completed a test transaction. It was that the system had begun to make sense.

That distinction—between completing a digital procedure and understanding the system that gives the procedure meaning—is where I believe much of Malaysia’s next digitalisation challenge now sits.

Digital from the outside, fragmented from within

From a national perspective, Malaysia has made substantial progress in digitalisation. Digital infrastructure has expanded, payments have moved online, social commerce has reduced barriers to market entry, and software that was once affordable only to larger organisations is increasingly accessible to small firms. Government programmes have accelerated adoption, while regulatory changes increasingly require organisations to interact with formal digital systems.

When digitalisation is measured primarily through connectivity, access or adoption, much of this appears to be a success.

Yet KRI’s recent work complicates that picture. Its research on micro and small enterprises describes digital adoption as widespread but often shallow, with customer-facing tools such as payments, marketing and sales more commonly adopted than integrated back-office systems. Technologies are frequently introduced as standalone solutions or in response to immediate needs rather than as part of a coherent organisational architecture.

That pattern is familiar from the organisations I have worked with. A business may accept QR payments but lack a consistent reconciliation process. It may sell through several digital channels while keeping its authoritative inventory count somewhere else. It may possess accounting software while important decisions still depend upon information held in the owner’s memory. Customer commitments may sit inside WhatsApp conversations, supplier details in another person’s phone, prices in spreadsheets, purchases in paper records and formal financial entries inside an accounting platform.

Each tool may function correctly. The organisation as a whole may nevertheless struggle to produce a reliable picture of what is happening.

This is a problem of fragmentation rather than simple exclusion. An organisation can be digitally connected without being digitally integrated. Indeed, digitalisation can sometimes add another layer to an already fragmented organisation. The paper process does not disappear when software is introduced; it may simply be joined by a spreadsheet, a messaging platform, an app and a new compliance portal. Instead of consolidating organisational knowledge, technology can distribute it across even more locations.

The business becomes more digital while becoming more informationally dispersed.

When the workflow does not resemble the software

I encountered another version of the same problem through a manufacturing business, where digital systems had to contend with the messier realities of physical operations. In that setting, questions about software were inseparable from purchasing, materials, inventory, production requirements, customer commitments, costing and the movement of information between people responsible for different parts of the operation.

A transaction inside software is clean. A real operating process rarely is.

A purchasing decision may begin because someone notices that a material is running low. The urgency of the purchase may depend on what is currently being produced, what has already been promised to a customer, what quantity remains physically available and whether a supplier can deliver in time. The financial system may see a purchase order, but the operational reality contains a chain of judgments made before that document exists.

The same is true of inventory. A system can display a quantity, but that figure becomes useful only if the organisation has established what counts as stock, when movement is recorded, who is responsible for recording it, what happens to damaged or unusable material and how physical reality is reconciled with the number on the screen.

Technology can standardise these movements only after the organisation has made them sufficiently explicit.

This is why implementation cannot simply force an organisation to behave like the software. The software has to represent the operational logic closely enough that people can recognise their own work inside it. Where the formal system and lived workflow diverge too sharply, workers create parallel processes, workarounds and unofficial records because the work still has to continue.

Seen from this perspective, resistance to a system is not always resistance to change. Sometimes it is information about a mismatch between the organisation imagined by the software and the organisation that actually exists.

The organisation that has to remember itself

The same question appears differently inside a membership-based institution.

Such an organisation has members, committees, programmes, payments, documents, vendors, inventories, minutes, financial records and recurring governance obligations. It also changes leadership, and this matters because knowledge can become concentrated inside whoever happens to hold a position during a particular term.

A member record may have one life inside a spreadsheet and another in a messaging group. A payment may be known to the treasurer but not yet reflected elsewhere. A programme may generate invoices, attendance records, photographs, vendor communications and decisions that later have to be reconstructed when somebody asks what happened. Minutes record formal decisions, but they rarely capture every operational assumption that allowed those decisions to be implemented.

Then the committee changes.

A new office-bearer inherits the title, but not automatically the institutional memory that came with the previous holder.

This is where digital systems reveal another purpose beyond efficiency. They are not merely tools for accelerating transactions. They can become repositories of organisational memory, provided the organisation has developed the discipline to use them that way.

The implementation work in this setting therefore became much broader than teaching people how to navigate software. Membership, user access, finance and payments, inventory, customer and vendor records, purchasing, selling and monthly operating routines all had to be understood as parts of the same institutional system. The practical question was not simply whether each module worked. It was whether the organisation could eventually understand itself through the information it was creating.

The challenge of digitalisation here was therefore inseparable from governance. Who records? Who approves? Who can see? What survives a change in office? What becomes authoritative when different records conflict?

These are technological questions only in part. They are also questions about institutional memory, accountability and organisational design.

What does it mean to be digitally capable?

The term “digital adoption” therefore risks collapsing several distinct conditions into one. A business may have digital access, meaning that it possesses a device, connectivity and relevant software. It may develop functional literacy, allowing users to execute a defined task. It may acquire conceptual understanding, in which users understand what the system is doing and why. Beyond this, it may develop operational integration, embedding the technology into routine organisational practice rather than maintaining it as a parallel activity. At a more mature stage, the organisation may develop adaptive capability: the ability to interpret information, diagnose problems and alter processes when circumstances change.

These are not equivalent achievements, nor do they necessarily develop automatically from one another.

The business owner in Miri did not lack commercial knowledge. A production team does not lack knowledge simply because part of that knowledge is embodied in practice rather than captured in a database. An association does not lack institutional knowledge simply because much of it remains dispersed among office-bearers, files, messages and memories.

What these organisations often lack is a mechanism for translating tacit and fragmented knowledge into a form that can be shared, preserved, interrogated and acted upon collectively.

This matters because policy interventions can easily diagnose such situations as deficits located entirely within users. The conclusion then becomes that SMEs need more digital literacy, more training or greater motivation to adopt technology.

Sometimes they do.

But another possibility is that institutional systems demand forms of knowledge, formalisation and abstraction that smaller organisations have never previously needed to develop.

The distinction changes where responsibility is located. If the problem is simply low digital literacy, the obvious intervention is more instruction. If the problem lies in the interaction between literacy, organisational practice, institutional design and implementation, training alone is unlikely to be sufficient.

The business before the software

Technology-adoption research frequently identifies complexity, perceived usefulness, compatibility, organisational readiness and employee knowledge as important variables. These concepts are useful, but ethnographic observation provides a different way of understanding what they mean inside the daily life of an organisation.

Compatibility, for example, may refer to whether a system suits a particular business model. At the level of lived practice, compatibility also concerns whether the categories embedded in the software resemble the categories through which users understand their work. Complexity may refer to the number of steps required to complete a transaction, but it can also describe the cognitive distance between an embodied practice and the formal abstraction through which a digital system represents that practice.

This is why a digital-transformation project should not begin with the software.

It should begin with the organisation.

A useful first question is deceptively simple: show me how this currently works.

Follow an order from the moment it arrives until payment is collected. Follow a purchase from recognition of need to supplier payment. Follow a member application from initial contact to formal registration. Follow a committee decision until somebody actually carries it out.

When the path is traced carefully, an organisation’s real information architecture begins to emerge. It is often quite different from the organisation chart and sometimes different again from what users initially describe.

A person may say that records are “in the system” until a particular exception reveals that the authoritative information actually lives elsewhere. A formal workflow may appear clear until a staff member explains the phone call that must happen before the workflow can proceed. An approval process may appear procedural until one discovers that everyone waits for the judgment of a particular individual.

These are not minor details. They are the organisation.

Reflexivity: I am inside the system too

Writing about “low digital capability” creates a methodological and ethical difficulty for someone who designs, implements and teaches digital systems. The language itself can reproduce the hierarchy under examination. The professional arrives with the software, the SME is described as behind, the professional possesses knowledge and the user is framed as lacking it.

My own professional position places me squarely inside that relationship. I began my career in policy and ethnographic research and later moved into organisational transformation, technology platforms, business analysis and SME implementation. I now lead a company that develops systems, training and operational support for smaller organisations. I therefore have both an intellectual and commercial interest in believing that better-designed systems can solve organisational problems.

That position cannot be treated as neutral.

When an implementation struggles, it is easy for a technologist to say that users resisted change. A trainer can say participants lacked digital literacy. A policymaker can say firms failed to take advantage of available programmes. Management can say staff refused to follow process.

Each explanation may contain some truth. Each can also protect the intervention from examining itself.

My earlier experience conducting ethnographic research with urban-poor youth taught me to begin with observation before prescription. That instinct became unexpectedly relevant again in digital implementation work. The settings were different, but the methodological lesson remained remarkably similar: the categories brought by the researcher or practitioner are not necessarily the categories through which participants experience their own lives.

The experience in Miri sharpened that lesson. Instead of beginning with the question, “What does this person fail to understand?”, I increasingly find it more productive to ask, “What knowledge does this system assume they already possess?”

That change in perspective does not require romanticising informality. Manual processes can generate serious weaknesses. Information held in one person’s memory creates risk, poor records constrain financing and succession, and undocumented processes become fragile as organisations grow. Yet recognising these weaknesses does not require treating existing practices as irrational.

An ethnographic approach begins by asking what those practices accomplish for the people who use them. Only then can we determine what must change, what can be retained, and what may be lost as formal systems are introduced.

The missing layer between adoption and productivity

KRI’s recent findings suggest that Malaysia’s digitalisation challenge increasingly lies beyond simple access. Digital tools are already present in many micro and small firms, but weak integration limits their effects on productivity, resilience and scale. Financing constraints, skills, informality and limited access to advisory support all contribute, while fragmented programmes can make the support ecosystem itself difficult to navigate.

I would describe the missing layer between adoption and productivity as operational capability.

Operational capability is the ability of an organisation to make its working processes sufficiently explicit to be represented within a system, maintain reliable information as part of everyday work, understand how information moves across functions, interpret the data produced, preserve knowledge beyond particular individuals and identify when the system and reality have diverged. It also includes the capacity to adapt when regulation, technology or market conditions change.

This capability cannot simply be purchased as software, nor can it reliably be produced through a single training programme. It develops through repeated interaction among people, routines, technologies and institutions.

This may help explain why digitalisation programmes can produce disappointing outcomes even when the technology itself functions correctly. A firm may receive software before it has clarified the process that the software is supposed to represent. Training may teach functionality without supporting organisational redesign. A consultant may configure a technically sound system while employees continue to operate through parallel informal processes. A government platform may be accessible from the perspective of interface design while remaining conceptually difficult for the smallest firms to navigate independently.

In each case, adoption has occurred.

Capability has not necessarily followed.

Who is missing from the dataset?

There is also a methodological implication that deserves more attention. If policymakers want to understand the businesses furthest from digital maturity, conventional digital research methods may contain their own blind spots.

An online survey about digitalisation is most easily reached by people who are already online. A digital maturity assessment requires a respondent comfortable enough with formal organisational categories to complete it. A webinar advertised through digital channels will attract firms already connected to those channels. A grant application designed to support digitalisation may itself require formal documentation, financial records and administrative confidence.

This raises an important empirical question: are the firms most in need of digitalisation support also the least visible within the mechanisms used to study and support digitalisation?

If so, the selection problem matters. Firms appearing most frequently in surveys and programmes may systematically differ from organisations that remain informal, digitally hesitant, geographically isolated or institutionally disconnected.

Quantitative measurement remains indispensable. Malaysia needs better and more timely firm-level data. Yet some questions require researchers to enter the organisation before expecting the organisation to enter the dataset.

An ethnographic research design can begin with questions that appear almost mundane. Show me what happens when a customer places an order. Where is the information written down? Who else needs to know? What happens when that person is absent? How do you know whether you made money from the sale? Where do you look when the numbers do not match? What happens to knowledge when the treasurer resigns, the storekeeper leaves or the founder is unavailable?

Such questions reveal an organisation’s information architecture without requiring respondents to know that the concept exists.

They also reveal the places where formal systems and lived systems fail to align.

From user-centred design to institutional reflection

User-centred design has long argued that systems should be built around the realities of users rather than requiring users to conform unnecessarily to assumptions embedded by designers. A similar principle deserves greater attention in public digitalisation policy.

The relevant question is not only whether a user can navigate an interface. Policymakers and designers should also ask what kind of organisation the interface assumes exists behind that user.

Does the system assume formal bookkeeping? Does it assume stable product classifications? Does it assume familiarity with taxation terminology? Does it assume continuous internet access, distinct administrative roles or documented organisational knowledge? Does it assume the person entering the information is not simultaneously the owner, salesperson, purchaser and accountant?

For associations, does it assume institutional knowledge survives leadership turnover automatically? Does it assume the member database, financial system, event records and committee decisions already refer to the same organisational reality?

For manufacturers, does it assume the physical workflow has already been standardised before it enters the system?

For many small organisations, these assumptions do not hold neatly.

Designing for such realities does not mean lowering standards indefinitely. Reliable records matter. Formalisation matters. Tax compliance matters. Data quality matters. The question is how organisations are enabled to move from one mode of working to another without mistaking the existence of a platform for the existence of capability.

A digital system can potentially function not only as an endpoint of formalisation but as a scaffold through which formal capability develops. Achieving that requires deliberate institutional and implementation design.

What policy could do differently

If Malaysia wants micro and small enterprise digitalisation to generate deeper productivity gains, policy should begin by distinguishing adoption from capability. A business using several applications is not necessarily more digitally mature than one using fewer but better-integrated systems. Measurement frameworks should therefore examine whether information moves reliably between functions, whether records are used in actual decisions, whether key workflows remain dependent on individual memory and whether users can resolve routine system problems independently.

Support should also begin with diagnosis rather than prescription. Before introducing technology, implementers need to understand how the organisation currently works, where information is created, where it is lost and which problems matter most to the organisation. For very small firms, this does not require sophisticated process engineering. Following the movement of a sale, purchase, payment or membership record through the organisation can reveal more than a generic digital-readiness checklist.

Malaysia may also need to treat implementation support as part of digital infrastructure rather than as an optional service added after technology deployment. Small organisations often require trusted intermediaries who can translate between regulation, technology and everyday practice. Such intermediaries might include associations, cooperatives, local business networks, training providers or community-based support structures. Their role is not merely to teach users how to operate software but to help organisations translate existing practices into formal, sustainable systems.

A stronger policy framework would also recognise progression. Not every firm requires an enterprise system, and not every organisation should be pushed toward the same level of complexity at the same time. The more useful question is not “What technology has this organisation adopted?” but “What organisational capability does this organisation need to develop next?”

Finally, digitalisation research should deliberately seek out organisations that are least likely to appear through conventional digital recruitment channels. Field-based qualitative work, longitudinal case studies and partnerships with trusted local intermediaries can reveal the realities of firms that remain invisible in national dashboards. Failed or abandoned digitalisation efforts may be particularly informative because they expose assumptions that successful cases conceal.

The ledger is not the enemy

It is tempting to narrate digital transformation as movement from an inferior past toward a technologically superior future. The paper ledger becomes a symbol of backwardness and the integrated platform a symbol of progress.

The organisations I have encountered complicate that story.

Paper can conceal inefficiency, but it can also embody an entire working logic that has developed through years of practice. A spreadsheet can be fragile, but it can also represent the first moment an organisation makes previously tacit information visible. A messaging thread is a poor institutional archive, but it may also be the channel through which an organisation actually coordinates itself.

Digital systems can create visibility, but they can also generate new dependencies when users learn procedures without understanding the structures beneath them. An organisation may become more technologically sophisticated while becoming less capable of functioning when the technology changes or when the person who understands it leaves.

The goal should therefore not be to replace paper with software as quickly as possible, nor should policy preserve weak processes simply because they are familiar.

The more demanding objective is to help organisations make their knowledge visible, transferable and usable.

Technology can play a powerful role in that process.

But technology is not itself the capability.

Conclusion: from compliance to comprehension

The woman in Miri eventually completed the task on the screen. If the programme were evaluated only through successful task completion, that might have been enough.

For me, the more important moment was subtler. The fields on the screen had begun to connect to something she already knew. The transaction was no longer simply a sequence imposed by an unfamiliar system. It had acquired a reason.

Manufacturing taught me a related lesson: software cannot create coherence merely by imposing structure on a workflow it does not understand. Association work revealed another: information systems become meaningful when they allow an institution to remember itself beyond the individuals temporarily carrying its knowledge.

Across these settings, the underlying question is the same.

What happens between adoption and understanding?

Malaysia’s next phase of digitalisation should not be judged only by how many organisations gain access to technology or how many applications they adopt. The deeper question is whether organisations develop the capability to understand, integrate and exercise agency through those systems.

Digital adoption tells us whether technology has entered the organisation.

Digital capability asks whether the organisation has learned how to make the technology part of its own way of knowing, remembering and deciding.

For Malaysia’s smallest organisations, that difference may determine whether digital transformation becomes another requirement they must comply with or a genuine source of productivity, resilience and economic agency.

Methodological Note

The Miri vignette is reconstructed from my professional experience facilitating SME digital-adoption training rather than from contemporaneous ethnographic field notes. I therefore do not present reconstructed participant speech as verbatim quotation.

The manufacturing observations draw on my direct professional exposure to PROMAT ESM, while the membership-based institutional observations draw on my work with PENIAGAWATI Malaysia. In both cases, the discussion focuses on recurring organisational and implementation patterns rather than presenting either organisation as a formal case study or as representative evidence of Malaysian firms or associations generally.

My interpretation is shaped by my position as both a former ethnographic researcher and a current practitioner involved in organisational systems, training and technology implementation. That position gives me access to the practical settings discussed here, but it also creates potential biases, particularly a tendency to see formalisation and information systems as desirable solutions. The essay therefore treats practitioner observation as a source of hypotheses that require further empirical investigation rather than as a substitute for systematic evidence.

References

Khazanah Research Institute. From Survival to Scale: Digital Empowerment for Malaysia’s Micro and Small Enterprises. Kuala Lumpur: Khazanah Research Institute, 2026.

Sangosanya, T. A., Mohd Salleh, S. S. M., & Saraih, U. N. “Pathways to Digital Entrepreneurship Adoption among Malaysian SMEs: Analysing the Determinants and Technology Readiness Dynamics.” Journal of International Entrepreneurship, 2025.