Oprah Winfrey made analytics news recently when she tweeted her fans to encourage those with a Nielsen box to tune into her OWN (Oprah Winfrey Network). Those boxes are used to measure viewership, but such encouragement is a strict Nielsen no-no. Oprah duly apologized soon after the company requested she remove the tweet.
This is an intriguing situation for a couple of reasons. First and foremost, it seems a stretch that Oprah – who owns her own network – wouldn’t know basic Nielsen policies. The fact that Nielsen was spelled wrong (“Neilsen”) in the original message suggests Oprah’s Twitter may have been hacked, though she claimed responsibility. Or maybe she was taking part in a sophisticated analytics experiment in cross-channel audience tracking, seeing just how much a single Tweet could move the TV ratings needle. Oprah was even accused of being desperate about ratings for her new network.
But, whatever the true story, the real desperation is the ongoing lack of credibility measurement data seems to have. Oprah’s “Twittergate” simply illustrates a few of the core challenges faced by companies diving into digital marketing, social media and the wonderful world of analytics.
- Imperfect & Inconsistent Metrics: There remains great frustration and confusion across media about which audience measurement solutions to use and what various metrics mean to business bottom lines. Different ratings agencies have different methodologies for measuring page views, unique visits and other essential metrics. Similarly, the major site analytics packages can be implemented to track user behaviors and traffic patterns in unique ways. As a result, this situation mirrors the world of TV; while few in the industry believe that Nielsen’s numbers are foolproof, everyone more or less abides by them. As the Times article put it, they are the “currency of the industry.”
- Gaming the System: Oprah is a huge name and the fact that she so blatantly tried to juice her numbers shows how anyone, no matter how successful, under enough pressure can be compelled to cheat. In the digital world, the problem is even more widespread – likely because there is even more confusion about what the numbers mean. There is a great deal of market tension at work. Because different industry stakeholders have different incentives and objectives based on what they’re selling, it’s no surprise they share varying numbers. Put simply, sellers of advertising space prefer to see higher traffic numbers and prices, while buyers of advertising like the quality demos/traffic at lower prices. As a result, auto-generated page views and content categorization remain a big problem. Additionally, les scrupulous content publishers game the system in their own ways – like putting Lady Gaga in every headline or linking to whichever celebrity is trending hottest at the moment.
- Lack of Auditability: To accurately track return on their investments, CMOs and other digital marketers need faith in the numbers. Without such faith, it will be impossible to gain confidence in underlying performance and audience behavior data. That means optimizing campaign budgets, tweaking offers based on test results or remixing content becomes less like a science, and more like guesswork
Fortunately, after a great deal of pressure from the IAB, key measurement services agreed to be subjected to auditing by the Media Ratings Council (MRC). The MRC was established at the behest of the U.S. Congress in the 1960s to, among other things, ensure measurement services are valid, reliable and effective. Back in the day, I served on the MRC’s Internet Services Audit Committee as early ad platforms were subjected to audits and when comScore and Nielsen first agreed to be audited.
This was during the very first generation of online marketing and Web analytics, but the core issue – trusting the numbers – remains the same now as it was then. And it is only going to get more intense as marketers must learn to identify and track users across more channels and devices, and seek to embrace more sophisticated measures, around customer engagement and marketing channel attribution.
We don’t expect the industry to establish a gold standard for measurement or a universal set of definitions to be used across every delivery platform and by every ad network, analytics software package or rating agency. It’s critical that all stakeholders in digital marketing understand the broad parameters of measurement and how different sets and sources of metrics relate to each other, as well as the unique details and definitions of the analytics toolsets they use. Effective data governance is another part of the story. These are significant steps marketers can take to get past the old curse of not knowing which half of their budgets are working and which half is wasted.