As if responding to Bitterman's swipe at mobile, players in the industry seem determined to prove their growing scalability in a flurry of press releases shouting out their latest metrics. A lot of these numbers are very large and generally defined, so they have limited use. But they are indicative of milestones and a growing sense of scale.
Mobile ad network AdMob says in its latest metrics release that it served 1.59 billion ads in September. While distributed worldwide, about 42% of them (672 million) were in the U.S. alone. Ironically, other recent reports talk about the large amount of unsold inventory at some mobile sites, especially lesser-known endemic content brands. AdMob's numbers demonstrate at the very least the sheer volume of mobile Web traffic that is already kicking in, and perhaps how much of it is coursing through a lot of off-brand sites. AdMob made its bones aggregating inventory from a lot of WAP destinations few of us ever heard of. But many of these sites, often community and personalized content portals, had generated a following among young audiences virally.
I have had several well-known media brands tell me recently that their newly launched WAP sites are already pulling down over 1 million page views routinely a month without marketing or momentum. AdMob's numbers suggest there is a lot more WAP activity than we might suspect.
And yet, raw metrics like this may be impressive on the face of it, but they don't tell us much about the nature of mobile Web activity. How many sites is this traffic spread across, for instance? How has the inventory increased in different content segments, and is this growth indicative of AdMob's increased appeal, or a deeper penetration of WAP?
Another question entirely is how much mobile video activity really is going on. MobiTV announced today that it surpassed 3 million subscribers. It announced 2 million subs back in February, so take-up is accelerating. MobiTV powers SprintTV, which is the likely source of much of the business. I have to say that I am not sure what that number really means. I guess it shows that at some point 3 million people signed up for whatever added charge Sprint or MobiTV requires. But because of the auto-renewal of these charges on phone bills, who knows how much actual use is going on? How much churn is there?
A more useful metric comes from MyWaves, the off-deck video portal. That site reported last month that it had reached 1.25 million unique users. Of course, those 1.25 million uniques may be churning month to month as well, but we are getting closer to a measure of steady video use. MyWaves also says it should surpass 4 million registered users soon. Registration numbers are always inflated, unreliable estimates of a real user base, but if we compare the registration to the usage logs, it suggests almost a third of sign-ups are coming back each month. MyWaves also said that time spent per average visit was an astonishing 20 minutes. Clearly these are people who really like the video experience (at least enough to drain their battery for it) and find it satisfying even on a small screen. The next question, of course, is whether this is an early adopter trait or something that can proliferate.
Mobile metrics are in the same place as Web and streaming metrics were in the late '90s and early 2000s. We reassure ourselves about growth and a coming critical mass by counting the highest number we can. Anyone remember the totally meaningless reference to site "hits" circa 1996? I recall those regular Arbitron raw counts of the number of people who had ever accessed a video or audio stream. They seemed enormous and promising at the time. Wow, this Web multimedia craze was really ramping up, some claimed in 1999. But too often they were counting every dabbler who tried to click a jerky video once, saw how crappy the experience was, and never tried again for another year or two. We were still years off from broadband approaching critical mass.
And so, too, no matter how many fat numbers we throw at ourselves about mobile media, the medium arrives the same way film, radio, TV print and the Internet did -- when it occupies a ritualized place in our lives. The real measure of a medium is not the number of people who dabble in it -- but the number who regularly, habitually depend on it for data.