Last week, Andreas Goeldi, the Chief Technology Officer at Pixability, shared some new data at the ÜberTube 2014 Brand Summit. In a presentation entitled, “By the Numbers: YouTube Science and Measurement,” he provided examples of how “Big Data” can turn conventional wisdom on its head. Pixability’s proprietary software scans 25 million keywords and one million videos per hour to provide brand marketers with Big Data about YouTube publishing behavior, traffic patterns, trending channels, content performance, key influencers, user sentiment, advertising patterns, demographics, and more. So, I grabbed a seat in the front row and took lots of notes.
Top 10 Branded YouTube Channels ONLY Have 270 million Views
Goeldi began by asking, “How big is big?” And he answered by saying the minimum number of views needed to get into the top 10 brand channels on YouTube was 270 million and the minimum number of subscribers was 700,000. Both lists include brands like the NBA
Goeldi looked at how this impacts content strategy. Do you need a viral video? Now, stop snickering. Some brand marketers see videos like “Volvo Trucks - The Epic Split feat. Van Damme (Live Test 6)” and want us to create one of those for them, too. But, how do you know it is really viral? You can promote your content with paid media. So, he said Pixability looks at the share of video views leading to social posts on Facebook and Twitter.
Using that yardstick, 4.9 percent of the 42.2 million people who watched “Pepsi MAX & Jeff Gordon Present: ‘Test Drive’” and then went on to share it. That’s better than the 4.2 percent of the 15.5 million people who watched “The Ohio State University Marching Band Performs their Hollywood Blockbuster Show” and then went on to share it.
Uh-huh. But how many people remember Van Damme’s epic split but forget Volvo Trucks? And how many people remember Jeff Gordon’s test drive but forget PepsiMAX? That’s the question that brand marketers should really be asking.
Successful Viral Videos Can Also Be Boring
And Geoldi had a surprising answer: “Successful videos are sometimes boring.” And he showed “Gmail’s new look,” which has more than 30.3 million views, “Apple - iPhone 5s - The new Touch ID fingerprint identity sensor,” which has over 3.6 million views, and “Samsung GALAXY S5 - Official Hands-on,” which has more than 4.3 million views. So, ask yourselves this question: Would you rather have people remember your video or remember your product?
What Video Content are Young Males Watching?
Geoldi used data to answer another question: “What do young males watch?” A major CPG brand thought 13- to 24-year-old males watched sports, lifestyle, technology & cars, and health & fitness videos on YouTube. What they actually watch is music videos, video games, and comedy – with all those other categories taking up less than a quarter of their viewing time. It’s 2014. Do you know where your children are?
Geoldi tackled another old chestnut: “Do YouTube viewers only like short videos?” He contrasted the videos that consumer electronics brands produce, which are generally 1-3 minutes long, with what people actually watch, which are typically 3-10 minutes long. So, how long should a YouTube video be? Long enough to reach a point.
Geoldi looked at another common question: “How often should you publish new content?” And he showed that make-up demonstratorMichelle Phan publishes new content about twice a week. Comedian Grace Helbig publishes new content five days a week. And Swedish video game commentator Felix Arvid Ulf Kjellberg, better known by his online alias PewDiePie, publishes new content six or seven days a week 51 weeks a year. So, the typical brand needs to create more content – from two to seven times more content – if it wants to keep up with the YouTube stars.
Brands also need to consistently create a wider variety of YouTube content to cover every stage of the buying funnel. Yes, they need to create awareness, but they can also explain products, build credibility, provide help, and encourage referrals. Unless, of course, their marketing slogan is: “I’m feeling lucky.”
Video Advertising: How to Target an Audience
Next, Goeldi took a quick look at advertising to this audience. He compared and contrasted the 30-second ad on TV with the many different types of ads – before videos, beside videos, in search results – of any length on YouTube. Yes, I know that Henry Ford once said, “Any customer can have a car painted any color that he wants so long as it is black.”But ad agencies probably shouldn’t tell their clients, “You can have any format as long as it’s a 30-second spot.”
He also compared and contrasted how you can (kind of) target your customers on TV with the (exact) targeting options on YouTube. I am beginning to see a pattern.
He compared and contrasted what you can measure on TV and YouTube. Any questions?
Unusual Video Advertising Can Perform Better
Goeldi then let this cat out of the bag: Unusual ad types can perform better. If you compare and contrast website visit length (page views), then a 30-second commercial will give you 100%, a 90-second tutorial will give you 118%, and 2.5-minute tutorial will give you 130%. Oh, did I mention that, unlike TV, there is no additional cost for running longer videos on YouTube?
Goeldi concluded the first half of his presentation with this observation: The ad campaign is the new focus group. Qualitative focus groups became popular in the 1960s – back during the Mad Men era. They gave brand marketers and their ad agencies more insights than any of the big three television networks could before spending big bucks launching a new ad campaign.
Now, there are lots of internet marketers and video content producers who still remember the old days –before YouTube Analytics was announced in November 2011 and even before YouTube Insight was introduced in March 2008 – when all we had was gut instinct and view counts to guide our decisions. But, although some of us may like to think of ourselves as modern versions of Don Draper, we should remember that most of us were just hungry entrepreneurs and starving artists.
Today, we have access to more YouTube data than we can shake a stick at. So, it’s probably time to change the way we create ads. Frankly, we should reverse the process: Conduct small tests on a wide variety of ads, find out which ones work best, and then ramp up our big ad campaign.
But wait, there’s more!
Goeldi also shared some new data in his presentation about Discoverability and Engagement. And he tied it all together. But that’s another column for another day. Stay tuned. Don’t touch that dial.