Think about the last time you utilized a search engine on the web to gain information on a specific topic. Regardless of the engine you favor, the results of a typical search can return hundreds or thousands of web pages from the billions out there in the webiverse. Yet most users seldom get past the first page or two before clicking through on what they deem to be the most relevant results. We seem to have faith that both our search terms and the logic behind the search engine combine to give us relevant and usable results right off the top… literally. What is behind that? To start, as the searcher we usually have a pretty good idea of what we are trying to find out about, and some experience at tightening up search terms to optimize our chances of finding it. The search engines have to utilize pretty complicated algorithms in order to determine how to list the huge number of results. If they do that well, you are satisfied and build your trust in their service and technology. If they offer you irrelevant results or results dotted with SEO content farms, then you just might switch engines, and that is bad for ad rates.
Some of the magic behind search engine results is a methodology for ranking pages. According to Google, they use “over 200 signals, including PageRank, to order websites, and (they) update these algorithms on a weekly basis”. PageRank essentially assigns a score to web pages based on many parameters, including how many other pages link to it. It is part of the Google competitive advantage, and is undoubtedly refined on an ongoing basis. Over at Microsoft, Bing reports that “the Bing ranking algorithm analyzes factors such as web page content, the number and quality of websites that link to your pages, and the relevance of your website’s content to keywords. Site ranks change as we review the factors that make up the ranking.” And at Yahoo! Search, results are ranked “according to their relevance to a particular query by analyzing the web page text, title, and description as well as its source, associated links, and other unique document characteristics.” Each has their way of ranking results to provide the best possible user experience, and it is critical to their mission.
Now turn you imagination to a media buyer preparing to make a DOOH ad buy for a product that will appeal to a targeted mix of geographic and demographic segments. Like the web searcher, you are skilled in defining what you are seeking, but not completely aware of all the potentially attractive destinations for your ad buy. How do we ensure that the ad buyer comes up with relevant results? We must be mindful of the fact that incomplete or irrelevant results will result in a no-buy or an ineffective buy scenario, both horrific for the industry. The answer will be an evolution toward a DOOH analog of web search page ranking algorithms. It has to happen, because there are too many networks to understand, and too many variables within those networks to analyze. We must make it easy for the buyer to be satisfied, just as search engines do for web users. It amounts to relevance or death.
At DSE 2011, we may have gotten a clue as to how that is going to come about. In the innocuously-titled session, “Top 10 Ways to Attract Advertisers to Your Network”, rVue CEO Jason Kates laid out a list of network quality indicators that might influence advertisers. (Full disclosure: our software integrates with the rVue demand side platform.) At face value, the top 10 list alone would be valuable to attendees and food for thought, but Kates decided to reveal a little secret: they also form the foundation of a network ranking algorithm for rVue. A murmur went through the audience, and pencils went to work. This might be important.
Consider the challenge inherent in rVue’s business strategy. They have well over 100 networks represented on their platform and a similar number of agencies using the tool to plan DOOH buys. When a buyer does a network search across any number of potential parameters, the tool must provide search results. Remember the principle of relevance or death: in these early days of automated DOOH planning, the results matter, as do the results of the results. They really have to get it right to retain buyers. So network search results must be ranked, lest they rely upon a random walk theory. But how?
The presentation document is available from rVue, but here are the top 10:
- Location: Targeted vertical niches are enhanced by the horizontal reach of quality locations.
- Quality: We are talking displays, media players, digital signage platform software and operations. They all matter, don’t skimp.
- 3rd Party Research: Independent validation of your traffic, impressions and even compliance.
- Content: Important enough not to get lumped under quality. It is either relevant and engaging or it isn’t.
- Audio & Video: Having both means a higher value than just having one.
- Full Screen vs. “L-Box”: Full screen is more valuable to advertisers. Hallelujah. Eventually, people will figure this out.
- Internet Connected: Amazingly, there are still sneaker-net networks out there. Go to the bottom of the results, please.
- Audience Linked to Transactions: Networks that exist in proximity to advertised products and services provide measurable results.
- Daily Analytics: Sounds basic to me, but apparently not everyone gets it or can deliver it. See item #2/software.
- Reputation: You are who you say you are, you deliver on promises, you act ethically. Not to be taken for granted.
How they score networks and come up with a ranking is rVue’s secret sauce, but clearly an important part of their value proposition to ad buyers. On the network side, just as webmasters have learned to explore and exploit page rank algorithms, operators can take it upon themselves to proactively enhance their ranking, and the exercise will help their positioning with advertisers whether or not the networks are engaged with rVue. Often times, that may require new investment in infrastructure, operations, content or outside services. Possibly a painful thought, but I think at this point we’ve all learned that if you just build it, they may not come. Networks must build and operate with quality, and they must demand quality from their vendors, service providers and staff. Today, I imagine it is relatively easy to rank networks using the 10 factors as a starting point. As advertiser demand for quality drives standards of operation higher, it may become more difficult. That will be a good problem for rVue and others to have to tackle.