Business Models in Transition: What Technology Really Changes About Companies

Technology first changes the economics of an activity

Many business models rest on a simple constraint: some activities are expensive. Another customer requires more service capacity, higher revenue demands more infrastructure, a new region needs distribution, and professional services often scale only when more qualified human time becomes available.

Technology can loosen that relationship. Software can be executed almost indefinitely once developed, data can be reused across multiple processes, and digital platforms can coordinate vast volumes of transactions without the organisation growing at the same rate. Artificial intelligence is extending this principle into activities that previously depended on human interpretation, writing or decision support.

The shift is visible in investment data. The OECD reports that real investment in software and databases across OECD economies has almost tripled since 2007, while digital investment more broadly has grown substantially faster than many traditional capital categories.^[1][2]

This is more than an IT upgrade. When the cost of an economically relevant activity falls, the set of services that can be offered profitably, the scale a company can reach and the resources required to reach it can all change.

1. Scalability begins with reproducibility

A company does not scale simply because demand rises. It scales when it can serve additional demand without cost and organisational complexity increasing in the same proportion.

Information technology can alter exactly that relationship. An NBER study using US firm data describes a mechanism through which IT allows efficient routines to be replicated across locations. In the firms studied, stronger IT investment was associated with sales growing faster than employment, a pattern the authors describe as “scale without mass”.^[3]

For a business model, the important issue is not whether a process becomes digital, but which part of the value proposition becomes reproducible. A consulting company may still rely on individual conversations while systematising research, data preparation, documentation and standard analysis. A retailer can expand assortment without adding equivalent floor space. A financial institution can handle standard transactions digitally and reserve human capacity for complex cases.

Two companies can therefore adopt the same technology and produce very different economic outcomes. One accelerates an existing process by ten percent; the other redesigns the process so that a meaningful share of its former marginal cost disappears.

2. The boundary of the firm becomes more fluid

Technology also changes what a company needs to own. Many business models used to be more capital-intensive because firms had to build their own servers, communications infrastructure, software, payment systems or administrative capacity before they could scale.

Cloud infrastructure, software-as-a-service, digital payments and specialist platforms turn a growing share of that infrastructure into services that can be purchased on demand. Some fixed costs become variable costs, and capabilities that were once largely confined to large organisations become available to much smaller firms.

This does not create unlimited independence. A company that owns less infrastructure becomes more dependent on platforms, APIs, cloud providers and technical standards. The strategic problem is not eliminated; it moves.

The regulatory debate illustrates how central this infrastructure has become. In June 2026, the European Commission informed Amazon and Microsoft of its preliminary view that AWS and Azure should be designated as gatekeepers under the Digital Markets Act because of their role as important gateways between businesses and customers. This was a preliminary regulatory position, not a final determination.^[8]

Technology therefore does not simply make companies more independent. It changes what they depend on and which dependencies have to be managed strategically.

3. Information becomes a factor of production rather than a by-product

Traditional companies often produced information as a by-product of operating. Sales were recorded, customer contacts stored, inventory counted and invoices archived. Digital companies can turn the same information into an active factor of production.

A sales system learns which customer groups respond to particular offers. Production data can signal when equipment is likely to fail. Usage data influences product development, while pricing and demand information can be analysed almost in real time.

Artificial intelligence extends that mechanism because more unstructured information can now be processed economically. Text, conversations, images and documents do not have to be fully converted into structured tables before they become useful.

A feedback loop emerges: the company delivers a service, observes how it is used, processes the resulting information and adjusts product, distribution or operations. The shorter and more precise that loop becomes, the faster the organisation can adapt.

Competitive advantage then lies not only in having the better product. It may also come from recognising earlier what should change in the product, price or process.

4. Distribution becomes part of the business model

One of the deepest changes in the digital economy has been the reorganisation of distribution. The internet reduced the cost of reaching customers directly, search engines structured attention, social platforms opened new communication channels and marketplaces aggregated demand.

This created business models whose economic core lies less in producing one specific service than in controlling an interface. Platforms coordinate distinct groups of participants and create value by making their interactions easier, safer or more efficient.

Research on two-sided platforms explains why these models behave differently from conventional linear businesses. A platform may rationally serve one user group very cheaply, or even below its immediate marginal cost, when participation by that group increases the value offered to another.^[4]

That can also change who the customer actually is. The person using a service need not be the party generating most of the revenue.

The trade-off is strategic. A company that gives a platform control over customer access gains reach and infrastructure, but gives up some control over pricing, data, discoverability and the direct customer relationship.

5. Technology changes who pays — and what they pay for

Technology changes monetisation as well as distribution. Software made subscriptions attractive across many industries, cloud services normalised usage-based pricing, freemium separated user acquisition from monetisation, and advertising financed services that appeared free to the immediate user.

Artificial intelligence may push this logic further. When software no longer merely provides functionality but performs parts of a workflow or produces concrete outputs, the conventional per-seat licence becomes less obvious. Providers can instead charge per case, transaction, unit of compute or delivered outcome.

Pricing then moves closer to economic output. At the same time, risk shifts between customer and supplier. Under a conventional licence the customer carries much of the utilisation risk; under outcome-based pricing the provider assumes a greater share.

A change in pricing architecture is therefore not cosmetic. It can alter which customers are attractive, how revenue scales, which costs the provider carries and how economic risk is distributed.

6. AI makes the difference between digitisation and transformation visible

The current development of artificial intelligence shows especially clearly why technical adoption and economic transformation are not the same thing.

The Stanford AI Index 2026 reports that 88 percent of surveyed organisations use AI in at least one business function. At the same time, AI-agent deployment remained in the single digits across nearly all business functions.^[5]

The gap between use and transformation is even more striking. A June 2026 study by the World Economic Forum and Kearney reports that only about 25 percent of companies say AI is already having a transformative impact. Many organisations still layer AI onto existing processes rather than redesigning work, decisions and business models around new capabilities.^[6]

McKinsey reached a similar conclusion in September 2026. Only 13 percent of the organisations in its sample were placed in the highest maturity group, where roles, workflows and the operating model are substantially reinvented. That group was far more likely to report meaningful enterprise value from AI than organisations that mainly provided general-purpose tools.^[7]

The pattern is not new. New technology first appears as a better tool. Only later do companies learn to build organisations that operate without the old technical constraint.

7. A business model changes only when an economic relationship changes

Not every digital upgrade is therefore a transformation. Replacing a paper form with a digital interface may improve a process without changing the economic logic of the company. Giving an employee an AI assistant can raise productivity while leaving product, distribution, pricing and cost structure largely intact.

Business-model change becomes more plausible when a fundamental relationship is reorganised: when additional customers can be served much more cheaply, when a capability that once had to be internal can be bought externally, when data itself becomes part of value creation, when direct customer access appears, when payment mechanisms or risk allocation change, or when a service becomes viable that could not have survived under the previous cost structure.

These shifts often interact. Cloud infrastructure can reduce capital requirements and accelerate market entry; better data can influence both analysis and pricing; a digital sales channel can lower transaction costs while simultaneously creating a direct customer relationship.

The relevant measure is therefore not how digital the company looks. It is whether its economic architecture has changed.

8. Technology can scale bad economics as well as good economics

Technological leverage has an uncomfortable property: it does not amplify only strong systems. It can make bad processes, poor incentives and unprofitable business models faster and larger as well.

A digital acquisition channel can accelerate customer growth, but if every new customer costs more over time than they contribute, the company is merely scaling its losses. An AI system can produce offers faster but cannot repair an unattractive value proposition. Automation can reduce handling time while spreading errors through a badly designed process at greater speed.

High scalability does not eliminate real costs either. Modern AI services require substantial compute, data infrastructure and operational capacity, so the economic bottleneck may move from labour to compute, integration and governance.

Technology does not abolish economics. It changes the parameters — and therefore the places where a business model can fail.

9. New dependencies replace old ones

Every technological simplification can create a new strategic dependency. A company that no longer operates its own servers becomes more dependent on cloud providers. A business that acquires customers through a marketplace becomes more sensitive to that marketplace's rules and fees. A company that embeds external AI models into core processes becomes more exposed to model pricing, availability, data access and switching costs.

The question is therefore not whether a company should do everything itself. Complete self-sufficiency would often be inefficient. The real question is which dependencies are acceptable and which capabilities need to remain strategically controlled.

A strong business model uses external infrastructure where it enables scale without surrendering the elements that create bargaining power, customer access or differentiation.

Technology consequently changes the definition of core competence. Some capabilities become less strategic, while others — proprietary data, process knowledge, customer relationships or integration capability — become more important.

10. Competition moves to the system level

When new technologies become widely available, many of their early advantages disappear quickly. A company may benefit from a new application for a period, but once competitors can buy the same software, the tool itself rarely provides durable protection.

More resilient advantages come from combinations: data, processes, brand, customer relationship, organisation, infrastructure and learning capacity. Technology then becomes one component of a system that is harder to copy than an individual application.

That is the strategic meaning of business-model innovation. The important question is not whether a company can access the same models, cloud services or automation tools as its competitors. It is whether it can combine those building blocks into an economic architecture that learns faster, scales more cheaply or binds customers in a way that cannot be replicated simply by purchasing the same software.

The real change sits between technology and organisation

The most interesting question is therefore not which technology a company uses. It is which former constraint becomes less binding because of that technology.

Perhaps an activity becomes cheaper. Perhaps knowledge becomes reproducible for the first time. Perhaps information arrives faster, direct customer access becomes possible or a service that was once local can be offered globally.

Only then does the strategic question follow: what can now be organised differently than before?

Companies that merely place technology inside existing structures can become more efficient. Companies that recognise which economic boundaries have shifted can change something more fundamental: how they create value, how that value is distributed and how much of it they are able to capture.

That is where real business-model change begins.