Statistical Warfare: Opera and iPhone

newsguy 0 344 Views

"Opera Retakes Leadership from iPhone in Mobile Browser Market" reads the press release that arrived in my email this week. "Opera is the world's number one mobile browser, overtaking iPhone in May according to data from StatCounter Global Stats" it continues, which certainly peaked my interest. Not least because, according to another set of figures, the Net Applications' Market Share report, the iPhone commands more than 60 percent of the mobile browsing share.

So what is going on here? Well, Opera insists that according to the StatCounter statistics its browser scooped a worldwide market share of 24.6 percent compared to 22.3 percent for the iPhone. Certainly if you look at those StatCounter figures the market share ordering is clear enough: Opera, iPhone, Nokia, iTouch, BlackBerry, Sony PSP, Sony Ericsson, Openwave and Android. But I am old enough to appreciate that first impressions can be a little misleading occasionally. I suspect that this is just one such occasion, because surely it is failing to compare like with like?

Let me explain. Take the iPhone and the iTouch numbers and lump them together, after all they both use the Safari browser do they not, on a mobile device do they not, and you get a market leading share of some 37.2 percent compared to the Opera share of 24.6 percent. That would be a fairer way of presenting the numbers I reckon, especially as Opera runs on different devices with different operating system platforms.

"Opera began the year in number one slot but iPhone overtook it in February. May saw Opera regain the number one position," commented Aodhan Cullen, CEO and founder of StatCounter. "It will be fascinating to watch how this battle plays out over the year."

Sorry Aodhan, but no it won't unless you start actually comparing like with like. I know that there is but I for one am not interested in seeing skewed statistics to support either side in the argument.

Dani AI

Generated

— The short answer: the numbers aren’t necessarily contradictory; they’re measuring different things. Different web‑analytics providers use distinct samples and definitions (page hits vs unique users, raw counts vs geo‑weighted counts), so a single month’s “leader” can flip depending on which metric you pick. (infoworld.com)

Specifically: some trackers aggregate raw pageviews from a panel of sites (that skews toward heavy‑use pages and whatever sites run their script), while others count unique visitors and apply country weighting to better approximate global user population. That difference alone explains many of the big mismatches between services. Analytics also rely on user‑agent parsing to label “iPhone”, “iPod Touch”, “Opera Mini”, etc., and those labels aren’t perfect — devices may appear separately even when they run the same browser engine. Finally, certain mobile browsers (notably Opera Mini) use server‑side proxy/compression, which changes how many page requests are generated and how they’re recorded. (en.wikipedia.org)

Practical takeaway: decide what you actually want to measure (pageviews, unique users, or rendering‑engine share), then pick trackers whose methodology matches that goal and always report the source and date. If you want a reliable trend, triangulate across multiple providers rather than quoting a single press release — press shots from vendors often highlight the metric that favors them. When comparing “Safari on Apple devices” it’s fair to group iPhone and iPod Touch (and even Mobile Safari by engine) — but be explicit about that grouping so readers know you’re comparing like with like. (infoworld.com)

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.

Edit Preview