GREAT: Drake Morton
Advertising Forecasts Are For Suckers
by David Koretz , Thursday, August 21, 2008
The online advertising industry is ground zero for industry analysts who make ridiculous predictions to attract attention.
Worse, it is a vicious cycle.
The bigger the prediction, the more likely it is to be quoted by venture capitalists, reporters, industry pundits, and CEOs of start-ups looking to raise funding. The more you get quoted, the more reports you sell. So analysts are incentivized to produce ambitious, if fictitious, numbers.
Invariably, financial bubbles get created because people ignore fundamentals and start using questionable math to justify their own logic.
The rest of us sit below the ether trying to make sense of it all, while being told “we don’t get it.”
Maybe.
Or maybe every time a new business model gets created in the media industry, people make up a whole set of useless new metrics to measure it that never seem to track back to revenue. Or profit. Or common sense.
The mobile advertising market is the most absurd of them all. Forrester Research predicted the mobile advertising market will be less than $1 billion in 2012.
Then things got crazy.
In September 2007, the Kelsey Group estimated that U.S. mobile advertising would grow from $33.2 million to $1.4 billion in 2012. Then Strategy Analytics predicted the global market would grow to $14.4 billion in 2011. Not to be outdone, eMarketer estimated that the US market will be $6.5 billion by 2012, and $19 billion globally. Finally, ABI Research threw down the trump card and declared that mobile advertising would be a $19 billion global business a full year earlier, in 2011.
What a difference a year makes. Those same firms are now revising their online advertising forecasts down. Predicting huge growth, and then announcing a market change and revising the estimates down is a waste of time. It would be much faster to just predict more realistic growth in the first place. Except then you might not be able to sell a second report.
If the paid professionals have a 700% spread on forecasts only three years away, what hope do the rest of us have to accurately forecast a market?
The key to rising above the noise is to ignore often-touted, but meaningless statistics. Here are three rules for assessing a market:
1.   Value is NOT a function of cost. The fact that social networking inventory, for example, is cheaper than other advertising options does not inherently make it a good value. Value is measured as cost versus return. Perhaps it is cheaper because it does not perform as well. Many articles have been written about advertisers getting negative ROI from social networking despite $0.15 CPMs.
Social networking can be a great value, or an awful value, but cost has nothing to do with it. When you measure value, you need to demand a quantifiable ROI measurement.
Ultimately, a marketing channel will be measured by the ROI it can deliver. Advertisers will measure the profit they can generate, versus the money spent, and the amount of risk taken.
2.   The success of an industry is not measured in venture capital. Slide may have raised money at a high valuation, but so did Kozmo.com, PointCast, Flooz, Boo.com and countless others. Like any other industry, not all investors are smart, and even the smart ones make stupid decisions every once in a while.
A flood of venture capital into an industry should be a warning flag, not a sign of prosperity. Hot markets get over-funded, leaving a sea of second-tier companies floating around aimlessly. Online video and social networking generated huge returns for a few winners, but left an army of also-rans behind to die.
A venture capital investment should be taken for what it is: one person’s opinion. An opinion that statistically is wrong 70-90% of the time.
3.   Understand the “how.” All too often, these forecasts are made without understanding how the market is going to achieve the outsized growth. In the case of mobile advertising, analysts have layered assumption on top of assumption to get to their number. In doing so, they ignore one of the most basic rules of scientific experimentation: You can change only one variable at a time.
To arrive at a forecast, analysts must estimate the rate of mobile Internet adoption, consumers’ willingness to tolerate advertising on the phone, advertisers’ willingness to shift large percentages of their spend to an unproven channel, and publishers’ ability to develop new advertising formats that will effectively leverage the channel.
Any time you have this many variables, proceed with caution. If no one can explain the “how,” run.
Only two things about advertising forecasts make financial sense: Analysts are trying to sell research reports, and CEOs are trying to sell venture capitalists.
And neither makes a compelling argument.
(c) 2008 MediaPost Communications, 1140 Broadway, 4th Floor, New York, NY 10001
