Is TV Still Dominant Ad Platform?

For pretty much most of the last half century, TV has held a dominant position among media for its ability to deliver the largest mass audiences with the opportunity for multiple impressions over short periods of time (in ad terms, reach and frequency). Now, according to research by Simulmedia, TV ad campaigns for mass awareness-focused brands in the U.S. rarely reach more than 75% of their target audiences.  Furthermore, one-third of the target audience gets more than two-thirds of the ad impressions.  In other words, TV is not reliably or effectively delivering the large, general, national audiences to advertisers who want them.
The root of this problem is the huge increase in TV networks, stations, and programming.  In 1970, 85% of TV households in the U.S. could choose from 3 (or fewer) commercial channels.  Today, there are dozens of channels that reach 85% of U.S. TVHH, and more than 90% of TVHH subscribe to at least one multichannel TV delivery system that gives them access to 50-500 channels. For the major broadcast networks, top-rated programs today garner ratings and shares that would be considered abject failures in the 60s and 70s.  In many major markets, the top network is a channel that wasn't considered a major network ten years ago, and broadcasts in Spanish (Univision).
For advertisers who want a big, general, national reach, buying spots in broadcast network prime-time is still vital - they still deliver the largest national audiences.  However, that strategy is no longer sufficient - it misses more than two-thirds of the audience.
In a post on Online Spin, Dave Morgan suggests that that is just the start of the problem for "big-reach" ad campaigns.  Advertisers and agencies can no longer rely on a single delivery system - they need to put together a package of buys across channels and markets.  However, Mason argues, the critical tools needed to efficiently assemble a mix of ad buys that would optimize the reach and frequency of campaigns aren't there.  There are audience measurement tools, which have been improved to better handle the increase in channels and fragmentation.  Yet the industry estimates that at least 20% of TV viewing goes to networks and program sources that are either unmeasured or inadequately measured by the current system.  And, even if the measures were there, the structure for media planning and buying spots is optimized for broadcast TV.  Mason notes:

Changing processes that have been set in stone for so long is a very expensive proposition, and not an attractive one when most of the marketplace is pushing to reduce fees and costs, not increase cost and complexity.... Now, with the Internet, and its massive scale, interactivity, dynamic ad delivery, census-level measurements and web video, connected TV's beginning to threaten conventional linear TV, the long-held notion that TV advertising wasn't broke enough to fix probably doesn't hold up any more.

If you follow the trades, there's a lot of experimentation and pilot programs testing ways to better measure audience viewing and impact, consistently, across a wide variety of channels and viewing options.  Many show promise, but there's a lot more work to be done to get to the level of reliability and validity that advertisers and agencies really need.

Source - TV's Reach and Frequency Problem,  Online Spin