What Happened to Reach? Doesn’t Anyone Want to Talk About Online Reach/Frequency Anymore?


It’s late February 2006 and so far you haven’t had to give a “Why Online Advertising?” presentation to any of your clients. You’ve gotten a hefty increase in budget and have even been present at the “adult table” during communication strategy discussions. You’re feeling pretty grown up, aren’t you? In fact, online advertising dare I say is not even considered “new media” but an integral part of the mix. Just a few short years ago none of this was a reality and we were all looking for the silver bullet. For most, the measurement of online reach and frequency was that answer… that magical potion that could compare the Internet to other media, mature the industry, prove its value to your clients, and of course shift budgets online. After years of long winded panels, articles, and arguments about online R/F, what happened? Why have we shut our mouths as an industry?

Let’s take a minute to remember why developing valid reach and frequency models for the Internet is complex. Traditional media planners calculate the number of people that the media they buy reaches overall (reach) and how many opportunities each person in their target will get to see the ad (frequency). In other words, planners buy media impressions, not ad impressions. The Web is different. Since ad servers count the number of ads that reach a target, buys are based on ad impressions, not based on site traffic.

Because of this, online reach and frequency data will always be inherently different and will come from a marriage of data solutions, such as ad server data, log files, and browser-based (cookie) information. The biggest impediments to developing accurate online reach and frequency metrics are data-based: reach information for planning comes from panel-based research, while impression information for buying and reconciling comes from ad servers, which are silent about the demographics of those “users” who viewed an ad. We were ambitious and bright eyed though, and ready to figure it out.

During the years 2000 — 2004, there was an enormous industry-wide call to arms. Every major research and technology company was going to get involved, from NetRatings to Atlas DMT. Even the ARF (Advertising Research Foundation) issued guidelines by which reach/frequency on the Web was to be determined. One day, the chatter stopped.

Among online advertising professionals, two schools of thought compete when it comes to metrics for planning and measuring ads on the Internet. One school insists that the main way, if not the only way, the ad industry will accept the Internet among the mainstream media is by comparing apples to apples. The other school of thought explains that GRPs (Gross Rating Points) and the Reach/Frequency curve are based on projections and statistical modeling—but given its inherent trackability, the Internet should be able to move beyond projections and models to measure what’s actually going on.

In 2003, Forrester released a study that predicted 58% of all marketers would use Reach/Frequency to plan their online marketing spend. Have we hit that level of mass adoption? Although savvy clients are demanding more and more to see online GRPs and planning against reach goals, I would argue that it is still the acceptation rather than the rule. So what happened?

Have the tools we developed become accurate and widely accepted?

Yes and no. There are two widely accepted cross-media planning tools from the panel-based research providers: Nielsen//NetRatings and comScore, and the software manufacturers that allow access to this data, IMS and Telmar. They’ve each developed tools for online use that mimic Reach/Frequency runs for television. The IMS tool, called WebRF, makes use of Nielsen//NetRatings data while the Telmar tool accesses comScore Media Metrix data.


  1. I guess the answer comes in the form of a fairly simple question: Can anyone tell for certain what the reach of the Internet is? Reach in Television involved pre-determined areas of influence known as DMA’s. Even the networks were able to rely on Census data to target Households. How do you do that on the web? We might as well go back to using “Hits” as the online metric, it’s about as valid. We made it to the “adult table” because of our ability to track reactions at an infinite level. That will also be the reason why we’ll end up owning the adult table. Stay the course and do not let old media habits presume their way is better. It isn’t, it was just the closest they could get to predictive.

  2. As the person who wrote the ARF standards when I was chair of the ARF Online Media Council, I can talk to the R/F issue ad naseum. I personally stopped writing and speaking about it from a frustration standpoint. That we need accurate metrics goes without question. Traditional advertisers are always asking us what the Web schedule adds to their traditonal media schedule from a R/F standpoint. But the fact remains that NONE of the vendors in the space paid any attention to the ARF guidelines or white papers. NetRatings and comScore do not tie their data back to third party ad server impression data. They are very rough estimates and in most cases overstate reach. Most planners don’t even know this, they just blindly use the numbers, because they are there. Even Doubleclick, which licenses the IMS system to report out NetRatings data, has not attempted to merge the two datapoints, even though most planners think, because it comes from DC, it has their data in it. Much of the blame can be directed at investment that stopped after the bubble burst. But we have to get over that eventually and finish the job. The reality is that if we do not finish the job, new technologies will come along to solve the metrics problem and the companies that are passively collecting agency, site and advertiser dollars for incomplete data will be playing catch up rather than leading the pack.

Leave a Reply to David Smith Cancel reply

Please enter your comment!
Please enter your name here