Alex Iskold at ReadWriteWeb has a post on Semantic Search: “The Myth and Reality” that for the first time made me able to clearly express the benefits of semantic search succinctly. My comment on his post:
The value in semantic search is simply about creating ontologies that allow for results refinement that can deliver a much higher degree of relevance than currently possible with query refinement. This can/has huge value to users and advertisers.
I closed the comment with the following peeve I’ve had with much of the discussion around semantic search.
I would also caution people to not see semantic search engines as destination sites the way people now regard Google. This is a whole new ball game.
The power of semantic markup is real and it’s starting to be realized. Semantic web was bound to happen – it was thought about from the very inception of the web. Advances in web computing have now reached the tipping point and now there is little holding back the semantic web. Over the next two years it will emerge in various forms as an integral part of the intelligent web (that in my definition includes XML, JSON and any API). Best of all, the only change that will be felt by the public will be improved web experiences.
I first became interested in the advertising and marketing potential of a semantic web two years ago upon learning about microformats. I posted back then on the potential of semantics to create new marketing technologies. The ensuing two years have seen a great deal of VC investment and progress in this space and my interest has continued to grow.
I’ve been diving headfirst into RDF, OWL, SPARQL and the top down vs. bottom up debate. Truth be told, as much as I love multivariate testing and targeting I started RAMP to devote most of my time on work developing more advanced advertising products that can use semantic markup and APIs. In that regard I’ve been lucky to be work very closely the past few months and learn a ton from Eran Shir, Jon Aizen and the rest of the gang at Dapper, a company whose semantic technology is as good as it gets.
The marketing products being created (I call them Adplications and the folks at Dapper call them MashupAds) are revolutionary in intelligence, functionality and best of all performance. At their core they combine the best elements of search and contextual advertising. The ads we’ve launched over the past three months are seeing 200-500% improvement on every performance metric and interaction levels of more than 10 per user. Giant wins for both the brand marketers and performance marketers.
The reason these products work so well is:
1) Content: Semantic markup allows a deeper understanding of the page
2) Context: Semantic web is at its core about mashups – the weaving together of relevant information and resources
3) Intelligence: Ad servers can provide dynamic functionality through APIs, (unlike search ads that need to be served through Google and have little/no functionality)
4) Control: 300×250 display units allow a great deal of dynamic functionality with a decent user experience
Delivered through ad servers these applications allow you to tunnel through (query & search) databases through a dynamic and consistent delivery of refined content. This experience, more specifically the use of ontologies to amplify relevance, has tremendous power to move consumers through the funnel. Most importantly, these aren’t ads. They are tools. They are not annoying. They are useful and even helpful. The only thing “ad” about them is that you have to buy media to distribute them.
Something else as a marketer that I love about these products is the idea of turning the funnel on its head — taking your site to your audience instead of waiting for your audience to get to your site. That has never been attainable on the web for businesses. The closest idea was/is email. Yet, despite how intrusive email is, it’s still the most successful ROI in digital. All the work you’re doing for SEO, all the clicks you’re buying in Search, all the affiliate commissions, all the pay to install widgets…it is all to get people to come to you. Now, with a media buy of $1mm for 100mm quality impression you can get your site functionality on almost every major publisher and reach almost the entire online population for your key segments.
Let’s have some fun and play this out…
100mm impressions x .2 CTR (current avg.)
200k visitors x 5% CR
If your LTV or CPA is $100 you have break even on the $1mm ad buy! This formula was with a $10 CPM here. Needless to say, the ROI possibilities in some verticals are amazing.
Back to Alex’s post. He summarizes by saying:
Semantic search is an upcoming technology that has set the expectations way too high. We have all been misled into thinking that these technologies are here to dethrone Google by delivering better search results. Neither of those things is true.
I respectfully disagree here. Expectations for the technology are not too high. In fact I think they’re nowhere near high enough. Nor have we all been misled. No one working in semantics that I’ve talked to thinks of this as a battle to dethrone Google. That was just a good hook from media and PR to grab attention. In fact, this is far from a battle at all. This is a quest – a quest to fulfill the true potential of digital intelligence through better understanding of web data.
There’s so much more to be said about the changes ahead. Instead of writing about them the next few weeks I’ll be sharing some live examples of these applications at SMX Advanced on Tuesday and at WidgetWebExpo on June 16th.
Also if you have time you might find this interesting. I’ve had this clip on the RAMP homepage for a while but I thought I’d share it here – the Holy Father himself speaking about the potential of semantic web.
Leave a Reply