The landscape of technology in 2006 now seems impossibly quaint. Smartphones were on the cusp of something transformative, but not yet ubiquitous. The word “app,” as a distinct product category, had only just started to slip into mainstream use. For economists or statisticians interested in tracking this early moment—when mobile app studios began to emerge, before the gold rush—it’s necessary to work with tools that are, themselves, relics of an earlier logic. ISIC 6201, covering computer programming activities, offers a foundation, but also demands a degree of interpretation that may leave some uneasy.
To begin, one must gather all firms registered or classified under ISIC 6201. In most economies, this is a well-populated set: software houses, freelance coders, established IT consultancies, and a scattering of would-be app developers, all together in the same bucket. The lack of granularity is immediately apparent. There’s a kind of flattening—mobile app studios are buried under the same label as firms designing payroll systems or logistics software. But that’s the nature of the data. There’s no help for it.
The next task is to filter, as best as possible, for mobile app studios. In 2006, the majority of mobile application work focused on feature phones—think Java ME, Symbian, or BlackBerry OS. The now-familiar world of iOS and Android development hadn’t quite materialized. Studios working on mobile content, whether games, messaging, or productivity tools, often described themselves in language that overlapped with traditional software vendors. Key indicators can be found in archived websites, business descriptions, or early press releases: phrases like “mobile content,” “downloadable ringtones,” “Java games,” and occasionally, the explicit mention of collaboration with handset manufacturers or telecoms.
Differentiating these early studios from firms offering telephone-based services—IVR (interactive voice response), SMS gateways, or premium-rate information lines—requires a further level of scrutiny. Both groups might appear, on the surface, to be engaged in “mobile software.” The difference often comes down to technical language. App studios discussed client-side application design, installation procedures, device compatibility, and, less frequently, relationships with the nascent app marketplaces that some handset manufacturers were beginning to trial. Telephone service providers, on the other hand, focused more on connectivity, server infrastructure, call flow design, or carrier billing. It’s a subtle distinction, but for analysts patient enough to read between the lines, the clues are there.
Text mining can help, but is rarely enough on its own. Automated scans for “Symbian,” “BlackBerry,” “MIDlet,” “mobile Java,” or “handset applications” will flag a decent subset of studios, but false positives abound. Small-scale validation, by manually reviewing company descriptions or tracing project announcements, improves accuracy. It’s a tedious step, and there’s a temptation to skip it—one that should be resisted if the aim is clarity over volume.
One may also look for evidence of participation in industry events, hackathons, or early developer programs sponsored by handset manufacturers or mobile operators. In 2006, some of these events were still relatively niche, but they often left a public trace—media coverage, event programs, participant lists. Studios with a presence at these gatherings were more likely to be working at the frontier of mobile app development, rather than on legacy telephone-based solutions.
Another potential filter comes from patent filings and technical white papers. Although less common among the smallest studios, a patent for a new mobile interface or a publication describing a device-specific application architecture suggests genuine engagement with app development. Not every such signal will be captured under ISIC 6201, given the patchwork nature of self-reporting and classification, but these records enrich the picture.
The landscape is, in retrospect, more diverse than most official figures would suggest. Some firms pivoted quickly to smartphone development as the market opened up; others faded, their business model tied to a particular carrier or feature phone platform. The methods outlined above produce an estimate—a portrait, really, rather than a census—of the sector at its inflection point. It is a reminder that the boundaries of technology often shift just as they come into focus.
The imperfections in the data are more than a nuisance; they’re a kind of historical record, reflecting the uncertainty and experimentation of the time. Looking back, one finds not only the seeds of today’s app economy, but the ambiguities that made tracking it such a peculiar challenge. Perhaps some ambiguity is inevitable in the process, and perhaps that, too, should be counted among the findings.