The future of forecasting - Global News Analysis - Industry Overview

Telecommunications International, Sept, 2002 by Stephen McClelland

By rights, forecasters in telecom should be overloaded with work at the moment. Every sector is bedevilled with the need to improve predictions, but after the worst 18 months the industry has ever seen, their credibility--following earlier and seemingly unrealistic market projections--now looks a little tarnished. The outlook is probably less clear than ever, and some sectors remain almost completely uncharted. "The 3G business environment is the biggest business uncertainty ever faced by mankind," states Tomi Ahonen, a leading forecaster, 3G consultant and author of books 'm-Profits' and 'Services for UMTS'. Strong stuff indeed, but if uncertainty suggests the need for intelligent prediction, how can forecasts now be believed when so many got so much wrong?

"The value of forecasting is not in knowing exactly what the future is going to be, it is rather in helping to prepare for the future," counters Ahonen. "A good forecaster knows the limitations of his forecasts, as well as how accurate the forecasts can be in any given forecasting environment." And structural change in telecom sectors such as mobile exacerbate this situation, according to Ahonen: "In mobile telecoms, I would say all forecasters had predicted that the time from around 2001 to about 2003-2004 would be exceptionally difficult to forecast precisely because of the fundamental changes taking place in mobile, such as mobile data services and awaiting the launch of 3G. We are now seeing it to be so."

Being proved right about the unpredictability of forecasting or a transitional glitch may get the experts off the hook, but forecasters such as Ahonen are unapologetic about the current 3G experience. "When considering the past forecasts for 3G--such as launches, subscriber numbers, handset sales, ARPU--I would say that practically all initial launch schedules have been pushed back, typically, by about a year. The forecast has been rescheduled, but not significantly altered. Most forecasts for 3G suggest that the biggest initial revenues will be voice, then messaging, but what we are gaining today is the increased accuracy in verifying the assumptions in the 3G forecasts, by drawing from early experiences."

But he acknowledges the uncertainty being faced and suggests the need to be more flexible. "The bigger the uncertainty, the more variety is needed in the forecasting tools. Certainly both quantitative and qualitative methods are needed. Quantitative tools give best accuracy with relatively stable environments where a lot of historical data is available. Considering the mobile telecoms future, quantitative tools work best to forecast only the short term, for voice and SMS traffic, revenues, and profits. For the more advanced mobile services, such as MMS, m-Commerce, mAd (mobile advertising), qualitative methods are needed. As real data emerges, more quantitative methods can be utilised."

Professionally, better practice seems on the way and a major non-profit network to encourage this--called the Foretel Forum--will be launched next year. The forum will seek to harmonise understanding and stimulate information exchange worldwide between specialists responsible for forecasting, market planning, modelling in operators, vendors, regulators, consultants and academics worldwide in fixed and mobile sectors.

Meanwhile, upcoming conferences will focus on developments in telecom forecasting, business modelling and market research: Telecom Forecasting Masterclass (25-27 September, London)--Ahonen will run a workshop on segmentation methods on Day 3 (http://ibctelecoms.com/forecasting); Telecoms Market Forecasting 2002 (7-10 October, Prague, http://www.iir conferences.com); Telecom Market Research (20-22 October, Esomar, http://esomar.nl/telecom); and Complete Telecom Forecasting and Business Modelling (6-10 April 2003, Oxford, England, http://www.virtualpredict.com). The Foretel Forum will be launched at this latter conference (http:// foretelforum.org) alongside a new publication, Journal of Telecom and Mobile Forecasting.