Three different questions behind one word
Calling a technology “relevant” might mean that companies already use it, that they have invested in related assets, or that analysts and executives expect it to become more important. These are not interchangeable indicators: each answers a different question and may describe a different stage of technological change.
Observed adoption requires defining who uses which technology and during what period. Capital investment, by contrast, approximates the resources accumulated in assets that may enable its use. A trend list usually selects technologies for their strategic potential, use cases or novelty, but does not by itself demonstrate widespread deployment.
For a company weighing an investment, distinguishing these measures helps avoid two mistakes: treating a forecast as proof of an established market, or concluding that a technology is unimportant because its current use is still limited. The relevant signal depends on the decision: sizing operations, comparing sectors or anticipating capabilities calls for different data.
What the technology-capital indicator measures
A CaixaBank Research analysis published on February 13, 2020, approximated the presence of new technologies in Spain’s economy using four EU KLEMS capital categories: software and databases, research and development, computers, and telecommunications equipment. The combined measure was expressed as euros of capital per hour worked, in constant 2010 euros. (https://www.caixabankresearch.com/es/economia-y-mercados/actividad-y-crecimiento/relevancia-tendencias-y-diferencias-tecnologias-del)
This approach allows comparisons across countries and sectors of the capital associated with those categories relative to the work performed. It can also be used to study accumulation trends and sectoral differences. It measures capital stock, not the number of companies using a particular technology or the returns they obtain.
The main caveat is that the categories are broad: they include software and databases, research and development, computers, and telecommunications equipment. They do not isolate AI or identify how much of the measured capital is associated with any specific application.
Business use: a more direct but bounded measure
Spain’s National Statistics Institute (INE) Survey on ICT and e-commerce use in enterprises asks about technology use and defines different reference periods: ICT-use variables refer to the first quarter of the survey year, while other data, such as e-commerce, refer to the previous year. The survey follows Eurostat methodological recommendations. (https://www.ine.es/dyngs/INEbase/es/operacion.htm?c=estadistica_C&cid=1254736176743&menu=ultiDatos&idp=1254735576799)
The latest publication listed on the INE page reports that, in the first quarter of 2026, 33.6% of businesses with ten or more employees used artificial intelligence and 39.4% used paid cloud computing services. For businesses with fewer than ten employees, the reported AI-use figure is 21.36%. The figures refer to different company-size groups, and the survey specifies the reference period for its indicators. (https://www.ine.es/dyngs/INEbase/es/operacion.htm?c=estadistica_C&cid=1254736176743&menu=ultiDatos&idp=1254735576799)
These figures describe reported use among the businesses covered by the survey. They provide a measure of reported uptake, but should be interpreted in relation to the survey’s population, technology definitions and reference period.
Strategic trends are not deployment statistics
A September 2022 Computerworld article summarised six trends that Gartner associated with the early development of metaverse technologies: gaming, digital humans, virtual worlds, shared experiences, tokenised assets and spatial computing. The article described adoption as early and fragmented, with large-scale adoption more than ten years away. (https://www.computerworld.es/article/2112967/gartner-esboza-seis-tendencias-que-impulsan-la-adopcion-temprana-de-las-tecnologias-del-metaverso.html)
That example shows why a trend should be read in light of its date, scope and qualifications. A forecast about potential uses or future time frames does not say what share of Spanish companies has already adopted them. The article concerned the metaverse specifically and estimates reported in 2022; it is not a general inventory of business priorities for 2026.
The same applies to lists of emerging technologies. Their value lies in organising hypotheses, opportunities and areas worth monitoring, not in replacing use surveys or investment data. A forecast can inform experiments or planning, but it should remain distinct from observed outcomes.
How to connect indicators without conflating them
Metrics can be read as complementary parts of a process, not as different versions of a single figure. Technology capital approximates accumulated resources; surveys describe reported use within a defined population and period; trend reports suggest technologies or applications that could gain importance. Together, they provide context, but they do not automatically form a maturity scale.
For an SME, a practical approach is to start by identifying the operational problem it wants to solve, then find the measure that fits the question. If the question is how many companies of a particular size use AI, an adoption statistic is appropriate. If it is how investment in technology assets is changing, a capital indicator is better suited. If the aim is to explore future applications, a trend report can serve as a map of possibilities, not as evidence of local demand.
A rigorous reading also accounts for dates. Categories, populations and reference periods can vary; the INE survey, for example, specifies company size and the period covered by its indicators. Comparisons require matching the relevant groups and periods before attributing differences to faster or slower technological change.
Three ways to read technology change
The indicators address different questions; this comparison helps readers distinguish accumulated capital, reported business use and strategic forecasts.
| Capital stock | Broad technology-related assets per hour worked |
|---|---|
| Reported use | Technology use within a defined company group and period |
| Future possibilities | A planning signal, not a deployment statistic |
Capital categories are broad; survey results depend on their population and reference period; trend reports do not measure deployment.
Which conclusions are useful for decisions
The strongest conclusion is not that a technology appearing on a list will prevail, but that the gap between interest, investment and actual use is itself relevant information. If a company is considering an emerging tool, it can treat a forecast as a reason to investigate and adoption data as a market reference, without assuming that either proves the tool will be profitable in its own case.
To assess a specific investment, it is useful to define an outcome metric—such as processing time, errors or operating cost—and a comparison period before expanding deployment. These internal measures must be defined in relation to the process and its starting point. A market signal does not replace a business evaluation.
In short, measuring adoption means counting uses under a specific definition and scope; measuring technology capital means approximating resources associated with particular categories; and analysing trends means assessing possibilities and priorities. Keeping these dimensions separate helps prevent forecasts from being presented as facts and clarifies what information is needed for each business question.
