How Chinese consumers interact with online video advertising

The study shows an industry struggling with a wide range of problems including flawed frequency capping, repetition of bundled ad inventory, non-existent data to support cross-device pricing, and more.

In the 'Dive Video' study, OMD China, AdMaster and Mintel explored both watching behaviour and advertising effectiveness. It goes a step beyond other studies we've seen that mainly focus, at a rather superficial level, on time spent and penetration.

As video becomes the main entertainment diet across China’s demographics, it’s as important to understand not only consumers' motivations and attitudes, but also how effective the medium is.

The facts, as Campaign extracted from CNNIC, eMarketer and OMD:

• The number of online video viewers in China is likely to cross 500 million this year.
• Chinese netizens spend 80 minutes per day watching online video, which makes up 16 per cent of their leisure time.
• People discover content outside of online video platforms, with internet search (41 per cent) and word-of-mouth (37 per cent) the top sources.
• Thus, 69 per cent already know what they want to watch before watching.
• 41 per cent of viewers consistently use the same video platform as a habit. Tencent Video, Tudou, and Baidu Video have the least ads, but that did not stop Youku and iQiyi (22 per cent each) being favourites, because viewers "just got used to it". 
• Forecasts for total adspend on online video peg spending at US$3.8 billion (RMB25 billion) this year.
• Mobile ads outperform PC ones in driving purchase intent (13 per cent vs 8 per cent), especially among female audiences.

Challenges in the realm of video advertising

"We talk about online video a lot, but maybe we're not asking the right questions," said Bhasker Jaiswal, managing partner of business intelligence at OMD China.

In terms of advertising, data tracking is not robust or does not exist for video advertising effectiveness, Jaiswal pointed out. Vendors struggle to price online video inventory across devices because there is no data to support, and for advertisers, this leads to limited post-buys. It is understood that the major online video players in China tried to form a unified DMP in the past, but talks did not materialise.

In addition, content is not a differentiator for big online video players, just a hygiene factor, because viewers expect to find the same content on most of them. Why? The competition for exclusive rights to popular programmes has led to a "rotation" of consecutive seasons on different platforms. To secure ad revenue, online video platforms are monetising with a package of same-order ads in each episode, and in every episode, of the programmes.

Binge viewing of online videos now becomes annoying to the viewer with so many ad repetitions, in the case of Alessandro Pang, director of business intelligence at OMD China, who saw the same Pantene shampoo ad 24 times in one such eight-episode binge session. Each episode had three ad breaks, so you can do the math.

This raises the question: is programmatic buying, which is supposed to give marketers better control of frequency capping, a farce then? What online video platforms won't tell you, in public, is that too much frequency capping means less ad impressions.  

What they will point out in public is the tricky situation of third-party firms (the three main ones in China being AdMaster, Miaozhen, Nielsen) producing different numbers in their ad monitoring reports for the same campaign.

"Because we are all using different measurement standards," said Frankie Fu, VP of sales at Youku at the OMD event where the research was unveiled. The current practice of using CPM (cost-per-thousand impressions) to sell online video inventory exacerbates this problem, he said.

There are no straight answers, but OMD's director of business intelligence Jeanette Phang provided some immediate implications for advertisers.

"Sometimes less is more. Shorter ads (15s) achieve higher recall," she said. Brand recall is best at a frequency below 5x, and product recall at a frequency of between 5 and 10x.

As for online video platforms, one of her key suggestions was to develop flexible monetisation and pricing models to reflect consumer responses to devices, frequencies and ad lengths.