Tomorrow, October 1, is a big milestone for me in two ways. First of all, its the the ten year anniversary for this technology that I am most associated with, 3G mobile telecoms. The other milestone? Come back tomorrow to read about that. But lets talk 3G.
My last 'real job' was at Nokia heading Nokia's Global Business Consultancy department, part of the 3G Strategic Marketing unit (we were the first unit in the Nokia of that time, where our department was set across both the handsets and networks divisions of Nokia). With the work of my department, I soon became the 'poster child' for the business case of 3G, and most of my Nokia era press visibility was directly related to 3G services and their revenues. My first two books were written around these themes.
Today nobody claims anymore that 3G is a failure or 3G might fail. But in 2001, when I was out there selling the story of 3G and its revenues, there were plenty of doubters. As the dot com bubble burst, 3G almost was the baby lost with the bathwater, the mobile industry had to work hard to convince investors to stay with it, and that 3G would indeed come. Some famous analysts predicted that 3G would bankrupt the whole telecoms industry.
TEN YEARS AGO
So. October 1, exactly ten years ago tomorrow. NTT DoCoMo, of the same NTT that originally launched mobile telecoms in 1979 (four years before that famous Motorola phone was sold in Chicago on Ameritech's first North America based cellular telecoms service, this industry was not born out of the USA, it was born out of Japan). So NTT DoCoMo on October 1, 2001, launched 3G, under their brand, FOMA and on a slightly modified or pre-standard version of WCDMA, the most common form of 3G also called UMTS, and the technology that is part of the GSM evolution path.
3G was a VERY expensive proposition. It meant a total replacement of all cellular telecoms gear from 2G, and in case of services, a total upgrade or replacement of all service equipment from 'basic' 2G, early 2G before WAP and 2.5G services. The typical country like say a UK or France or Italy would need to expend about one Billion dollars per operator/carrier per country, to install the 3G network equipment, only on the cellular network (assuming the other network, services and billing parts were already upgraded from basic 2G to so-called 2.5G).
Then there were the 3G licenses, which varied very widely from totally free licenses like in Japan and Finland to Germany and the UK where one carrier/operator license cost roughly 7 Billion dollars each. For the first about 100 mobile operators/carriers to acquire 3G licenses, mostly in the Industrialized World, the average price was about 1 Billion dollars and this number was very widely reported (even though it was totally distorted by the UK and Germany, so the median license fee was somewhere near 200 million dollars, a very reasonable cost for a restricted market entry into a highly lucrative mobile telecoms business.
But this was a crazy gamble if for nothing else, that it involved unprecedented numbers, built on almost no facts or evidence of any commercially viable services. The early costs of 3G were estimated at 300 Billion dollars in equipment and licenses, cited commonly as the biggest infrastructure investment of all time (that number would balloon to more than twice by the time the decade was done, as more countries joined the 3G club). The Economist wrote on 13 October, 2001, in its cover story that the 3G mobile data project was the biggest gamble in business history.
Well, I was not deterred. I was the first person to explain in public how 3G operators would make their money and my 3G money story was the most widely reported story out of the GSM World Congress in 2001. I had not developed that story, it was developed at my consulting department at Nokia, by my smart analysts, econometric modellers and forecasters. But when the press came to call, or the major conference needed a Nokia speaker to explain 3G revenues, they sent of course the head of the department: me. And I told that story so many times, I knew it inside-out.
SO WHAT WAS THE VIEW BACK THEN?
I just did a trip down memory lane. In preparing for this blog story, I decided to go back and see what was the accepted 'common wisdom' about 3G, just prior to NTT DoCoMo launching it. So, the very last moment, before the first true commercial service launched, what did we, as an industry, think. Very revealing.. There was one body whose reports were THE most quoted and repeated numbers for the 3G opportunity: those of the UMTS Forum. This was not one analyst house, the UMTS Forum produced reports based on a consensus approach (usually not actual consensus but a majority agreement view, you'll never get technologists to agree totally on the future). So the UMTS Forum reports were always based on reviewing the current market views, forecasts and analysis, and then producing what was a kind of 'average' view, and then asked its panels to judge that view. Only if there was majority agreement, those points would be put into the major UMTS Forum forecasts about the future of 3G.
A very handy resource to look at now, ten years later, to remember exactly what was our industry view back in around September of 2001, about 3G (and the future of mobile, the internet, and telecoms).
REPORT NUMBER 13
The most quoted industry report about 3G (or any mobile telecoms in fact) in the year 2001, was the famous UMTS Forum Report Number 13. That was released in April of 2001. There was an update to that Report, the UMTS Forum Report Number 17 which came out just before the first launch of 3G, in August 2001, but that only touched on a couple of numbers and projections. And in Report 13, the writers mentioned several times, that the base case was the same as in the big UMTS Report Number 9, the first comprehensive 3G business case as expressed by the UMTS Forum from October of 2000.
For the most part, if we take Report Number 13, we make the few modifications from Report 17 that supercede Report 13, and we add the items from Report Number 9 that are the same but not repeated in Report 13, we have the best picture of the prevailing view of the mobile telecoms industry. And would you know it, the industry made its forecast for a 10 year period, from year 2000 to year 2010. Very convenient now, as we have end of year 2010 numbers, so we can compare the forecasts to the reality.
And remember, the longer you forecast into the future, the greater the chance of error. And remember, when these three reports were written, not one 3G network was in commercial production, not one subscriber was using any paid services on any 3G phones. None. Zero. Zip. But boldly they went, and forecasted a future on zero data, on a magnificent future, with only barely some early SMS data to go by, to try to imagine mobile data usage, and some very peculiar national-only Japan mobile internet data. There were no iPhones, the phones mostly did not have color screens, there were no cameraphones except in Japan. Music phones had not emerged, social networking was a concept almost nobody even understood at the time.
THE TECH ENVIRONMENT
Lets start outside mobile. The UMTS Forum Report 13 (and Report 9, both where modified by Report 17) made two relevant tech projections about non-mobile technologies. Internet users and fixed landline telecoms users. They projected that in year 2010, there would be 2 Billion internet users, and 1.5 Billion fixed landline users. The Internet user forecast is perfect. Exactly spot-on, 2 Billion. But the fixed landline user base did indeed grow, and followed their projection pretty well through year 2005, but then the strange thing happened, fixed landline subscriptions peaked and started to decline. They were down to 1.2 Billion by 2010. Note that this is the most difficult thing for a forecaster, to predict an inflection point, ie in this case, that the growth would turn into decline. Its easy to project a straight line from the past, to continue as a straight line into the future. The inflection is the challenge and this is where the report went wrong.
But that is pretty good, for a bunch of 3G mobile guys to predict the internet users exactly correctly, and while they were off by 20% from their projection for fixed landline - on a 10 year forecast, that is still a good result, as it still reflected tremendous growth during that period.
MOBILE INDUSTRY OUTSIDE OF 3G
Then lets get closer to 3G. What of other forecasts of the technologies related to mobile telecoms, which were not specific to 3G (but mostly would include 3G obviously). Here it gets interesting.
The UMTS Forum Report 13 and related documents forecasted that mobile subscriber numbers would grow dramatically in the period - by far the fastest growth of any technolgoy in fact. They predicted mobile subscribers would more than double in the ten year period, and reach 2.25 Billion subscriptions. A huge growth projection. The reality? They were massively off. They were off by more than 100%. Inspite of enormous growth, they still missed more than half of the real growth, which hit 5.1 Billion mobile phone subscriptions by end of 2010. Wow. So the industry expected that today, the globe would have a moblie phone subscription penetration rate of about 33% per capita, yes bigger than the internet even, and in reality the world at the end of 2010 had a penetration rate of ..74% mobile subscriptions per capita.
This is a good point right here, to understand exactly why it is, that year after year, the honest pundits will come back and say 'I got it wrong, the mobile industry grew more strongly than I expected'. Almost all analysts and experts felt the growth projections ten years ago at the limits of the plausible, and were looking for any signs to say that can't be. The safe bet seemed to be especially for analysts with long telecoms backgrounds (ie mostly from the fixed landline side, not mobile) and thus the biggest names in telecoms forecasting at that time, to make their own projections to be a bit more conservative, under the numbers of the UMTS Forum, which seemed the safer bet to them. It certainly could not be bigger, if you don't want to use the same number pick one a bit lower than this. And then, when those projections not only were met, but actually exceeded, year in and year out, then continuously we got those stories of 'we are near saturation now' and then next year, 'well, we are NOW near saturation' and again the next year 'but now we are certainly near saturation' and so it went. Eventually ever more of those pundits would notice it looks pretty silly to have to apologize for the mistakes, and they started to add more of an upside guess to their forecast and actually get closer to the truth (obviously I didn't need to do that, as I was armed with not just the best raw data but also the most comprehensively analyzed data - with the most realistic econometric model in existence - all obviously as work coming out of my department).
What else was in Report 13 about mobile telecoms overall? Ah, there was the total industry revenues projection. The UMTS Forum felt that there would be significant price erosion in the industry, but far more subscriber growth and traffic growth, and felt that the total revenues of hte industry would explode in the decade to come, growing 5-fold. They said that by 2010 the total mobile operator service revenues would be about 1.5 Trillion dollars (1,500 Billion dollars). Haha, the past decade did see mobile telecoms show the biggest growth of any industry, but not quite that big. They industry grew about 3-fold in that time, to 928 Billion dollars (error of 38%). As we've seen two recessions in the period, including the biggest economic crash since the Great Depression, I think we can forgive part of the error, and they did get the big story right, that mobile telecoms would see the biggest growth of any giant global industry on the planet.
Then what of our monthly phone bills? The UMTS Forum Report 13 projected that by 2010 the price of voice calls on mobile phones would fall so much, that our monthly voice call fees would be cut to about a third of what they were in year 2000. That by 2010, the world average voice calling monthly charges would be 11 US dollars. You know what? The real number was 11.13 dollars! They were off by one percent. That is perfection in forecasting! And incredible, to hit that number ten years into the future! There was no Skype, most markets hadn't heard of MVNO's and the number of carriers/operators was far smaller in most markets than it is today. Yet they hit that number on the head. Congrats.
But then on mobile data, they didn't get it quite right. The UMTS Forum felt that by 2010 the world average spending on non-voice data services would bring in a monthly payment across all mobile subscribers on average of 9 dollars. The reality was only 5.32 dollars, an error of 41%. Bearing in mind that mobile data was very rare in 2001 and apart from SMS, used only in some advanced markets like Finland and Japan, I think that is still not a bad result, I'd give it a fair grade. They were stabbing in the dark. Mobile data did grow very strongly in the past decade, but didn't reach quite the lofty levels they hoped. To put it another way, the UMTS Forum felt that voice vs data should have been 55/45 split by now. In reality it was 68/32. That is not a very bad error, but yes, it is significant error.
What is our biggest mobile data type? Messaging. They took the combined average price of SMS and MMS. How did the UMTS Forum see the price erosion of mobile messaging? They felt mobile messaging would decline strongly in price, roughly speaking down to one third what it used to be, and today the average message would be worth 3.5 cents. Reality? Close but no cigar. In the past decade mobile messaging price dropped to about a fourth, and was 2.6 cents (average of prices of all SMS and MMS sent) at the end of 2010.
They have been pretty good one could say. But as we read more, it gets more messy. They did get several significant factors very wrong. The UMTS Forum felt that 9% of voice calls would be videocalls. Less than 1% of them are on mobile. That is an error of 89%. They felt that mobile operators would earn 15 Billion dollars out of mobile advertising by now. In reality (there are VERY widely fluctuating numbers but a fair mid-point) the rough average total revenues from mobile advertising last year were about 6 Billion dollars, and far less than 2 Billion would have gone to the pockets of the operators. An error of 87%. And it gets even worse with mobile paymnets. The operators should have been earning 35 Billion dollars out of the sales commissions of mobile commerce and mobile banking and mobile payments. The reality of what operators took in was far less than 1 Billion dollars last year of those mobile money services. An error rate of 97%
There were more forecasts of individual service types, I think I stop here, as when we get to the more 'trivial' type of services, their relevance to the overall view is no longer that significant. But lets go to 3G.
The main point of Report 13 (and Report 17 and Report 9) was to map out the market opportunity out of 3G, not out of other mobile services and revenues. What did they say about 3G, that is the most significant point to consider today, as we celebrate 10 years of 3G. What did the industry honestly expect the world of 3G to look like, ten years later, before anyone had seen one 3G phone on one 3G network used by one paying 3G customer, anywhere?
The number of 3G subscribers was to be .. 630 million by end of 2010. Just to put this in context, that would be faster growth than any other technology outside of mobile, so faster than the internet, faster than TV, faster than credit cards, faster than radio, faster than computers, faster than anything in its first 10 years, except one other industry from mobile. Original 1G analog mobile did not grow anywhere near that fast, but 2G digital mobile did actually grow faster (passed 1.1 Billion in the first 10 years).
The other significant 3G forecast they made was 3G revenues. The UMTS Forum predicted that by 2010 the total 3G revenues would be a massive 320 Billion dollars - 3G alone would become bigger than all of mobile telecoms revenues at the time of the forecast being made in year 2001.
How did reality come to meet the forecast? Check it out. UMTS Forum said 3G subscriptions should be 630 million, the real number was 732 million! They made a bold strong upbeat forecast that would signal a very robust powerful new industry. They got it wrong, the reality was 16% better than they had foreseen. This is the kind of error you want from an industry consensus view. Especially on the biggest business gamble of all time.
And talking of business. Customers do not equal dollars. How did the revenue forecast pan out? UMTS Forum wanted 320 Billion dollars in revenues year 2010. The reality? Try 350 Billion dollars! Also the forecast was too conservative, the reality in money terms was 9% better than what many said was a rosy and optimistic view back then.
So 732 million people have not only a 3G phone, on a 3G network, but also have their subscription on a 3G pricing/data plan. And the money earned by the 3G industry alone now, is more than the total mobile industry service revenues were ten years ago. 3G makes so much money, those 'crippling' 3G licenses? They could be paid off - totally - out of the gross profit (EBIDTA) earned by the 3G industry just in one year now! And the 3G licenses run 20 years typically. How incredibly profitable will this 3G industry be?
Happy birthday 3G. I was out there in the troubled times in 2002 and 2003 when so many were saying 3G would kill the telecoms industry. I believed. I showed numbers from Japan, from South Korea, from early numbers in Europe etc. I believed. And I argued and argued. Slowly the phones came and the services and the subscribers and the revenues and profits. Today 3G is a huge success only growing bigger. Today all the talk is about 3.9G and 4G and beyond. But yes, ten years later nobody doubts 3G anymore. 732 million subscribers - 14% of all mobile subscriptions on the planet, bearing in mind most of those are now in the Emerging World countries where many live on one dollar per day - and a massive 350 Billion dollars of service revenues. Thats bigger than global television, boys and girls. Thats far bigger than the total internet revenues, including subscriptions, including advertising and service revenues. That is our 3G. And who wrote the first business book into the biggest gamble in business history? Yours truly! My second book, M-Profits has stood the test of time, into multiple printings, it was just translated into Chinese only last year. No rest for the wicked eh?
So, October 1, 2001, ten years for 3G. Thank you NTT DoCoMo for launching us into this adventure. All stats quoted from year 2001 were source UMTS Forum Report 13; stats for 2010 mobile subscribers, internet users and fixed landline users: ITU. All other stats from 2010, source: TomiAhonen Almanac 2011. You may freely quote this story, and any numbers from it. Please do help share this story and help celebrate 3G as the success it is, after all - this was the biggest business gamble of all time and we were witness to its improbable success. And tomorrow, come back to this blog for the other big anniversary also on October 1.