Why SMEs Aren’t Going As Digital As They Could Be

This article is based on the findings of a six-month qualitative study of small and medium enterprises across Georgia, conducted by Loialte with the financial support of GIZ (Deutsche Gesellschaft für Internationale Zusammenarbeit).

The report showed that the biggest barrier to digital transformation is not capital, infrastructure, or technology, but something narrower and harder to fix with another round of generic AI training.

The 21st century is considered the digital era. The world is becoming increasingly digital, and the transformation process is accelerating day by day. There is no field of human activity where modern technologies have not caused an evolutionary process. Business activity is no exception. On the contrary, in the 21st century, the success of a business largely depends on how well it keeps up with technological progress, therefore, the study of the digital transformation process of companies is becoming more and more relevant worldwide. In addition to the fact that monitoring the dynamics of companies’ digitization is very important, it is no less important to study the evolution of the term itself – digital transformation, because depending on the pace of technological development, this term carries a different load every year.

In addition to the fact that international companies such as the World Bank, OECD, and the European Commission systematically conduct studies on the digital transformation of companies, which are usually limited to publishing raw data, such as what percentage of companies use artificial intelligence or CRM (Customer Relationship Management) systems, additional studies are conducted in governments to identify factors hindering the digital transformation of companies.

In 2026, Loialte, with the financial support of GIZ, conducted a study on the digital transformation of small and medium-sized enterprises, which lasted for 6 months and included 54 companies from 5 different sectors. Given that the study was not quantitative, the goal was to surface patterns of digital transformation obstacles, not produce statistics.

What the field showed

One construction company participating in the study purchased a drone. The drone now completes in fifteen minutes what previously took a full day of surveying. Despite this successful experience, almost all other processes at the same company are still performed the same way they were ten years ago.  If you ask the owner why, the answer won’t be money. Drones weren’t cheap. The answer is that team members can’t determine where else digital technologies can simplify processes and reduce costs, or how cost-effective the investment will be.

A similar trend was repeatedly identified during the research process, which included in-depth interviews and five focus groups: in the fields of trade, manufacturing, tourism, construction, and agriculture. 

The biggest barrier to digital transformation is not financial. It’s awareness. And not awareness in the soft sense of “people don’t know AI exists.” Almost everyone knows AI exists. The gap is much narrower and much more practical: owners don’t know how to connect a specific tool to a specific process inside their own business. They can name ChatGPT. They can’t tell you what part of their daily process it should replace.

And if we look at the digital transformation process of small businesses through this prism, we will see that a large part of programs and projects is misdirected.

It’s not the money – it’s the lack of clarity

Ask any SME owner anywhere in the region why they haven’t gone digital and the first answer will almost always be money. Software is expensive. A website costs thousands. Proper CRM is unavailable. 

The deeper pattern, which was repeated across all sectors in the study, is that companies are reluctant to spend on digital tools because they don’t see a tangible benefit – businesses aren’t refusing to invest in something they understand and want – they’re refusing to invest in something they don’t understand.

This is important because “if the diagnosis is incorrect, the treatment is ineffective.” Put a grant into a system where companies don’t yet understand what to spend it on, and most of the money goes to the wrong tools, gets lost in the middle of the project, or isn’t used at all.

How might all this look in practice?

The contrasts inside a single sector tell the story better than any summary.

Trade Sector –  The study involved a pharmaceutical company that created internal automation, a warehouse management system, and a digital journal that simplifies the business management process. At the same time, another company involved in the study, with a strong social media presence, sits on piles of customer data that nobody knows how to use. Same sector. Same country. Different worlds  –  separated not by capital but by the question of whether anyone in the room knows what to do with the data they already have.

A tourism company said it has replaced several staff members with AI, integrated a CRM system, and is running a marketing campaign that would have required hiring a small advertising agency two years ago. Meanwhile, a hotel manager participating in the same study said, “We need someone to tell us specifically what to use and how to use it.” The tools exist – but not the bridge between the tool and the process.

The contrast is starkest in agriculture. Georgia produces qvevri (clay pot) wine – a method that is included in the UNESCO World Heritage List. The product is world-class. The digital marketing and online sales infrastructure around it is, in most cases, limited to a Facebook page. As the companies participating in the study stated, while the purchase of agricultural machinery is a priority for them, in similar circumstances, software subscription feels absurd by comparison. The financial barrier is real here. But behind it sits the same pattern: when no one has explained what online sales would do to the bottom line, the question of cost  never seriously asked

Compared to other sectors, the situation is different in the construction sector – companies, as they describe it, are in the so-called “comfort zone”. The business is successful, decision-makers do not understand why they should change what’s already working.” Drones are bought because they have a clear income. Everything else is waiting – not because of money, but because no one has put their position in the language that management actually speaks.

The Paradox of Artificial Intelligence

The finding that was repeated across all groups was this: everyone knows about artificial intelligence, almost no one has integrated it into a real business process.

The use of AI technology is mainly limited to three areas.

– Information retrieval;

– Generating generic text. Personal experiments, usually after work hours, usually by one curious employee.

– Systematic integration into the real work process – CRM, which generates subsequent reports, analyzes the financial process, detects anomalies, and manages the marketing process – and these operations take place without the intervention of the entire team of employees, although as mentioned above, only a few organizations involved in the study used this method;

The reason is not a lack of technical access. The point is that going from “I used ChatGPT once” to “Now AI does this specific thing every day at my company” requires,  somebody to sit down, look at the company’s actual processes, and decide which one to change first. That somebody usually doesn’t exist inside the company.

And the external trainings are mostly general in nature, not addressing specific issues. One notable exception was noted in the study – a tour operator that used artificial intelligence to process a database found that instead of two weeks of human labor, the process could be reduced to one day with the help of artificial intelligence.

An important detail is that this happened because the owner ran a ‘digital experiment”  that, as he himself noted, no one had taught him at trainings and seminars. No one taught him – That’s the real bottleneck – The proper translation layer between the technology and the business.

What actually helps

The research identified a trend that was noted in one form or another by representatives of almost every company in every sector – “not more training, or more grants, but more guidance  –  someone to come in, look at their particular business, and tell them what to do.” 

 This sounds modest. In political terms, it’s a significant shift.

Most existing support programs  –  training programs, awareness campaigns, general webinars  –  are built on volume – measured in “thousand of companies were trained on basics of AI issues… “

According to the study, this approach has reached its ceiling. Companies no longer need another general overview of “what is digital transformation”. They have “already seen these videos”. What the study revealed is the need for individual diagnostics and assessment of where they are in their digital transformation and what specific steps they can take, taking into account their size, sector and budget.

 The model that emerged from the research consists of three parts. First, individual digital diagnostics – short, structured, sector-specific assessments that create a prioritized action plan rather than a general recommendation. Second, awareness-raising in the language of business outcomes, not technical specifications – owners want to hear “this reduces errors in your warehouse by 30%” rather than “this is a cloud-based ERP module that is essential for running a company”. Third, longer engagement instead of one-off workshops – because choosing a tool is the easy part; implementing it, training the team and adapting the process is where most digital projects gradually stall.

This is not all exotic. Variations of it exist across the OECD, sometimes in the form of digital innovation hubs, sometimes in the form of consultancy time voucher schemes. The lesson is not that emerging economies need a fundamentally different model. The point is that the model exists, and the SME sector in the region is demanding it so loudly that someone needs to listen.

Why this matters beyond Georgia

The structure of the SME sector in the Eastern Partnership, the Western Balkans and parts of Central Asia is very similar across countries. Small companies, thin layers of management, owners working four jobs at once, training programs designed around averages rather than specific indicators. If the principle of “funding for awareness” is valid in Georgia, it can almost certainly be said to be valid in Armenia, Moldova, North Macedonia, Uzbekistan and a dozen other countries where the same grant programs are funded by the same donors and produce the same uneven results.

Which raises a question worth ending with.

If the real obstacles are practical, business-specific, process-level gaps in awareness and skills, then perhaps the question is not how much support is offered, but where in the “transformation journey” it arrives – and how well it will be designed around those specific gaps?

And one final note, in the spirit of the piece itself: AI was used in editing this article – ethically, as it should be wherevers it’s used. The analysis is the author’s own.