That Facebook and Pinterest are in the news isn’t news, but their latest turn in the spotlight provides clues about their aspirations to rise in search prominence by mining consumer intent. For Facebook, it’s the beta version of “M,” which The Wall Street Journal calls a “digital assistant” available within Facebook Messenger and contends could enable Facebook to garner more direct-response advertising dollars. In Pinterest’s case, it’s visual search: technology that enables Pinners to highlight objects and then search for similar objects, with obvious implications for search advertising.
In this Q&A, Mike Grehan, Acronym’s Managing Director and Chief Marketing Officer, offers his thoughts on these two new developments in the context of how search continues to evolve around consumer discovery, intent and recommendation. Coincidentally, just before Pinterest’s visual search technology became public, Grehan moderated a panel at Conductor’s annual C3 confab in Manhattan that included Vikram Bhaskaran, who heads strategic partnerships at Pinterest. Among other things, Bhaskaran noted that Pinterest now handles 4% of all searches in the United States and that 67% of all Pins are from brands.
Q: How do you think Facebook M will impact Google?
A: It’s probably more of a supplement to what you can get from Google because Google can’t crawl Facebook’s content.
Q: How does social search differ from what people get from traditional search engines?
A: Accessing content inside social search is like tapping into a network of trust. When I search Google, computer algorithms pull together a lot of content and rank it according to the algorithms, but it’s not always “right.” For instance, if a spammer wanted to create a problem with Google search by posting content about a virus children could catch and said the best way to get rid of the virus is for children to drink disinfectant, that could rank number one on Google. But if you go to social search, and you ask the same question about a virus, you’re tapping into a network of trust. The people you are talking to are parents, doctors, nurses and so on. So what you get back from this human algorithm is something more verifiable because it’s peer-group vetted. Most of the organic content that Google returns is not verified.
Q: So how will Facebook go after direct-response budgets?
A: Facebook is going to tap into intent exactly the same way other companies have. Given the amount of data it understands about its end user, if it can detect intent in searches, in exactly the same way Google can detect intent, Facebook can decide whether to serve you an ad or not. If I do a search on Google for “digital camera,” trying to infer my intent is difficult. Am I looking to buy one, sell one, whatever? But if I say, “buy digital camera,” then I’m going to see ads for that. I don’t think it’s a case of the attention ad being thrust in your face. It’s understanding the intent before you’ll see that ad on Facebook.
Q: So does this pose serious competition for Google?
A: Not really, because people don’t use Facebook as a general-purpose search engine in the same way that they don’t use Google as a social network. So I think it depends on the audience itself. It’s like saying is Twitter is a competitor to Google. For the advertising dollar, everyone has always been a competitor, the same as conventional press, radio and TV. My intent to buy a camera at Google might be different from seeking peer reviews for cameras at Facebook.
Q: How about Pinterest’s visual search? Is that a threat to Google?
A: Not in any way at all. The closest proximity there is to understand is that Pinterest is about images while Google is about text, even though Google has gotten pretty good at images. If I search on Google for images, I’ll get images from any old website that happens to have those images for whatever reason. The difference with search at Pinterest going forward is it’s about human-curated topics and concepts and not around keywords.
Q: According to the Journal, Pinterest considers its visual search a discovery engine. Is this accurate?
A: Microsoft beat them to the term “discovery engine” long ago when it rebranded Bing as a discovery engine.
Q: So where are these developments leading us in the evolution of search?
A: What we’re headed toward now in terms of search is beyond pure discovery. It’s what some researchers refer to now as “suggested discovery.” What that means is you search for one thing, topic or concept, and what you get is in much the same way as Amazon can say, “people who bought this also bought that.” Whether it’s at Google, Pinterest, Bing, Facebook or LinkedIn, you find things because of what they know about your peer group. You will see suggested discovery. It’s also creeping into TV apps at the moment: If you enjoyed this show you will enjoy that. Like Netflix does.
Q: So are Facebook and Pinterest recommendation engines?
A: There is a difference between recommendation, which is collaborative filtering, and suggested discovery. In collaborative filtering at Amazon, you might learn that people who watched Star Wars also enjoyed Star Trek. It wouldn’t say people who watch Star Trek mainly wear blue socks. That’s a different thing.
Q: Vikram Bhaskaran shared some interesting facts about Pinterest at C3. Which ones were most noteworthy to you?
A: Well, he said that from a marketing perspective, 67% of all Pins actually come from brands. That’s because marketers want to reach consumers when they are thinking of the future: What should my home look like? Where should I take my next vacation? What style should I wear? From a search technology standpoint, Pinterest believes human beings are better at answering those questions. He called it “human-powered search” and noted that search is 50% of the way people use Pinterest and the platform now handles 4% of all searches in the U.S. In this regard, he said one of the company’s main focuses is to reeducate marketers who have been trained to believe that Pinterest is strictly a social network.
