n the vortex of the economic downturn, advertising-supported media companies are facing not only dramatic spending decreases, but the massively disruptive effects of digital substitution. After an unprecedented three consecutive years of contraction in real advertising spending, advertising-supported media companies are asking when the turnaround will come, and wondering if advertising spending patterns will be forever changed.
Based on more than 1,000 hours of research into advertising spending across 12 sectors, a recent study by FTI Consulting forecasts a 13% decline in advertising spending in 2009 and another 1% decline in 2010. This follows decreases of 2% and 5% in 2007 and 2008 respectively. While the report predicts a small upturn in 2011, real advertising spending will not climb back to 2008 levels of $284 billion until late 2014.
Looking at the trend in spending patterns, traditional print media (particularly newspapers) will continue to bear the brunt of the spending decline, and will regain prior years’ market shares only in rare cases. Internet substitution has permanently changed print media’s market share. Over the coming years, the Internet’s share of advertising spending is expected to grow at a steady pace, rising from current levels of roughly 9% of total advertising spending to 15% in 2015. Exacerbating the problem for traditional media companies is that one dollar of advertising spending on traditional media is replaced by just 33 cents on Internet sources. Thus, the fastest-growing segment of advertising spending substitutes digital cents for each previous dollar spent.
The Impact of Online Advertising on Traditional Media
Our forecast research considered historical advertising spending patterns by medium to estimate the rate of growth of the Internet over the next five years, determined what percentage share will represent a steady state for the Internet and assessed its impact on traditional media.
The research indicated that television was the best analogy to the Internet in terms of its impact as a disruptive technology. The relationship between household penetration and advertising market share of TV is very similar to that of the Internet. After its introduction in 1949, television experienced rapid household penetration, reaching 50% of homes in four years and 90% in only 13 years. It captured about 15% of the advertising market during that time period, 20% after 25 years and eventually settled, 40 years later, at its current level of about 25%. Television put dramatic pressure on newspapers, which accounted for 37% of advertising spending in 1949 but only 29% of advertising spend by 1962. Radio’s share of advertising spending contracted by 50% over this time period.
When advertising levels will recover to 2008 levels of $284 billion
The Internet also took approximately four years to reach 50% household penetration. In comparison to TV’s 15% ad share in 13 years, Internet ad-spending share is expected to reach 10% by the end of 2009. This is 15 years after Internet usage debuted in the home and 11 years after broadband connections were introduced. FTI’s research suggests that the Internet will eventually peak and settle at approximately a 25% advertising market share. This forecast is based on a dynamic advertising-spend regression model, which uses the historical relationship between five variables to predict future ad spending: consumer spending, private investment, the unemployment rate, online market share and the incremental impact of either a U.S. presidential election or the Olympics.
Until 2000, real GDP growth and real advertising spending correlated quite directly. Advertising spending represented about 2% of GDP. When the economy grew, advertising spending grew and when GDP contracted, advertising spending contracted disproportionately. Since GDP is comprised of roughly 70% consumer spending, 10% private investment spending and 20% government spending, FTI found a stronger relationship between advertising spending and consumer outlays and private investment.
But spending and investment alone did not predict the trends experienced since 2000 quite as well. Therefore, to improve the accuracy of the model, particularly as GDP contracted, we also factored into our analysis three other variables.
First was the impact on consumer spending of employment, which we captured with changes in the unemployment rate. And, because of the effect of online price performance on total advertising spending and the resulting pricing pressures on traditional advertising media, we added a variable to reflect the change in online advertising market share. Finally, we captured the average boost in advertising spend during a U.S. presidential election and Olympic Games years.
The result of these additional factors was significant. The historical correlation from using our five factors was high, and the post-2000 correlation held even better.