PR News Q&A: Separating the ‘Real’ Data From the Flawed

Israel Mirsky

Israel Mirsky, executive VP of emerging media and technology at Porter Novelli, believes that communications pros shouldn’t take the word of metrics vendors on whether the data they are using is trustworthy. “In many cases, it’s not real…and all the analysis in the world won’t help you if the data you use is flawed.” In this Q&A, Mirsky touches on blending social media measurement with more traditional metrics, a topic he detailed further as a presenter on the panel “Media Measurement: From Mainstream to Social Media” at PR News' PR Measurement Conference on March 1, 2011, in Washington, D.C.

PR News: From your experience with clients, what do they generally miss about measurement?

Israel Mirsky: Clients tend to overlook two major things when it comes to online measurement. First, they tend to take the word of measurement vendors that the data that is used is representative or “real.” The truth is, in many cases it's incomplete in some or all channels—sometimes because of limitations of software, sometimes because of privacy considerations. The solution is testing, but most clients and agencies haven’t devised testing methods that actually measure the quality of data capture.

And second, survey data on the general population or target does not reflect social media, and social media data does not reflect the general population. They’re two different segments of the same overall group with dramatically different behavior patterns, which influence one another. Using social as representative of how people think is grossly misleading.

PR News: How do you blend social media measurement with more traditional PR metrics, or can you?

Mirsky: If the client wants impression metrics, blend the impression metrics you accrue from your online programs with your offline programs. Awareness and similar metrics are still meaningful—you just have to be clear about exactly who you’re measuring, and why.

PR News: Any unusual or unique metrics that you've implemented with clients?

We are very proud of our work to build PR and social media inputs for Marketing Mix Modeling calculations. As part of that project, we constructed algorithms that calculate the reach of influential posts—both owned and earned—for each of the social and offline channels we work on, with accuracy and granularity.

PR News: Which measurement's days are numbered?

Mirsky: I have a personal vendetta against the estimates for blog reach that many of us are forced to use because they are all we have. As an industry, as soon as we can move toward more meaningful reach estimates, we should, because on the other side we’re moving toward tracking on conversion events or bottom-line goals. Attempting to optimize your work against bottom-line goals based on bad data is more luck than science.

PR News: What is one best practice you'll offer that Measurement Conference attendees will take home with them?

Mirsky: Demand and scrutinize the Boolean search used for any analysis concerning your brands. Your analysts will hate you, but it’s important to agree on the nature of the net being cast. If you don’t understand how the Boolean search used to make up your conversation analysis was constructed, you can’t spot holes in the conversation you’re attempting to bring back—and you can’t ensure the data used is actually answering the question you’ve asked.