Not many people know this, but Zara is a very unique fashion company compared to other mega fashion brands. Instead of focusing on inventing the next clothing item people want to buy , Zara built an advanced technology that can bias or push their entire clothing line production in blazing speed to meet what real customers are actually buying across the stores. As Harvard published as case study, "This "fast fashion" system depends on a constant exchange of information throughout every part of Zara's supply chain" (http://hbswk.hbs.edu/archive/4652.html)
The fashion retailer has become one of the more successful fashion companies in the world – based on their unique way of manufacturing and distributing their clothes. As opposed to most fashion companies, Zara never manufactures lines of clothing and hope that the market will like it. Instead, they have highly efficient method for tracking what the market wants to buy, getting it into their factory, biasing their line of production accordingly, and distributing back to the stores. This behavior/technology differentiates them, as they're able to meet their marketers' needs rather than promoting fashion innovations. The combination of their biasing-technology with the real-world market, based on what is highly sold is truly fascinating and is one of the hallmarks of their success.
So, what does Zara's methods have to do with online video and discovery technologies, and what can you learn from it?
The reality is that video sites share similar needs, and that is the ability to not only create content users want to watch, but ideally to be able to bias users towards the content that is attractive to advertisers and has high demand. The reality is that there is no real scalable solution in place just yet and it's really hard to do. When such solutions are in place, the effect can be massively profitable for publishers as it was for Zara, and highly effective for advertisers who buy video inventory--the supply will be closer to the demand for specific types of premium content / verticals.
(1) Publishers offer much more video content on their site than they used to 5-7 years ago. They either produce it, buy it, rent it or syndicate it from someone. The result is a larger video inventory per site
(2) Video content is usually segmented into three main categories: UGC (such as YouTube's), Premium (such as Revision3'), Super Premium (such as ABC). As you climb up, CPMs are higher, and video views available to sell are smaller
(3) On most sites in the world, there are parts of the sites (i.e. categories such as Health, Technology) that are much better sold, at higher prices, and more desirable by advertisers than others. Problem is that generally publishers have less video views available on those categories to offer advertisers to buy. The demand is higher than the supply.
Result -> Publishers want more views in certain desirable categories. Users want to watch (only) interesting videos. Advertisers want to advertise on specific types of videos, and are willing to pay more for it
This post will discuss how video sites can offer users what they want to see while promoting what advertisers actually want to buy and are willing to pay a premium.
With Zara's case study, high CPMs and strong demand for specific categories in which it is difficult to increase streams, publishers should leverage anything they can to maximize good a discovery experience on their site for their users while being able to maximize revenue. Here are some ways to do it:
(1) Content is king
Before you declare "sold out" on the highly monetizable niche categories on your site, try to increase your video inventory on those sections. An expensive way to do it is to produce it yourself, or buy it. Other ways can include using embedded videos from YouTube and other sites that allow embedded videos. Most users don't recognize whether the player is the publishers, or a 3rd party site video player. They just want for content availability. This is something great sites like Huffington Post, ReelSEO, and others recognize and leverage.
(2) Become a promoter
Allocate places on your site that get meaningful traffic, usually the homepage or search page, and have 2-5 video links that are tailored to those niche sections. The conversion rate on those links will usually be below 1% but still can yield meaningful uplift in streams given the high traffic. For convenience, 0.5% on 30M page views on the homepage gets another ~2M streams you can sell annually on high CPMs.
(3) Use advanced biasing and discovery technologies
There are innovative ways of using mathematical algorithms to scientifically analyze video viewing patterns and predict what people want to watch in real-time. This type of technology is called "Video Discovery”. This is done by analyzing massive amounts of information being created as users watch videos and determining user's engagement (from video length preference, time spent on the site, day of week, Geo-location, how many videos viewed and multiple other metrics). Some of those technologies also come with a capability to seed meta-tags, and categories that publishes may want to bias and promote for business reasons and the technology will provide automatically. Usually publishers place those automatically generated discovery links on story pages, video pages, and even home pages.
The bottom line is that it's not enough to push users to watch more. It's important to join hands with advertiser's needs and put in place ways that can make everybody happy. Users and Advertisers align. A true win-win.
About Our Guest Expert: Adam Singolda
As the Founder and CEO of Taboola, Adam is leading the company business strategy, execution and business development. Taboola is the provider of the world's first content monetization vehicle through personalized video recommendation and highly targeted ads. Taboola guarantees that site visitors view more videos, stay longer and increase overall video revenue.Prior to founding Taboola, Adam served as an officer in a research elite unite in the Israeli National Security Agency (I-NSA) for seven years. Adam focus in the unit was data-mining and "unsolvable problems." Adam is an honored alum of the IDF's elite "Mamram" computer-science training program, and finished first in his class in the officer's academy of the IDF.
Follow Adam on Twitter @adamsingolda