This is a PowerPoint-free open Q&A session. It’s pretty freeform. I am going to try to put as much structure in as possible, but be warned — it’s pretty freeform. That said, there are some excellent jewels in here. If the freeform is too willynilly for you, please do read the bold spots.
Note: This whole Q&A session is (mostly) paraphrased. While the concepts came right from the horses mouth(s), the exact order of the words may have been shifted in translation from them to me.
Danny Sullivan (@dannysullivan) opens by saying “Hello, welcome to the SEO in 2014 session. SEO is Dead. Everyone go home. Just Kidding.” Danny is the moderator of this session (a perfect fit) and he’s full of jokes this morning. Before kicking it off with the first questions, Danny mentions that Matt Cutts isn’t here because “contractually Matt Cutts isn’t allowed to travel East of the Mississippi.” In his place we do have Brian White – the Program Manager for Search Quality and Webspam at Google – but with Not Provided as a hot agenda item it’s already shaping up to be quite the…say……….Google hot seat session.
We start with some kick off questions: how do you get 100% not provided?? Isn’t than an oxymoron?? What’s up with the Hummingbird agorithm shift that no one knew about until Google revealed it about a week ago??
Danny to Brian White (Program Manager, Search Quality, Webspam, Google —@BrianWhite)): “What can you tell us about the Not Provided shift; what happened? Will keywords always be not provided? Will we get Keywords back in another way?”
Brian White points to web security, saying (paraphrased): We’ve been working to protect our users for quite a while now. We made Gmail HTTPS only in 2008, then we switched to HTPS for signed-in users to protect our searcher, today we’ve arrived at a place where we’ve decided we need to implement Not Provided as a new measure to protect users.
We’re really concerned with helping users search securely. Not Provided really is about security and protection. This decision was not easy.
Webmaster Tools does provide rich keyword data. We’re hoping this can help people reach their goals.
Webmaster Tools is kind of a happy medium; it gives you some good keyword insight into keywords that are driving traffic to your website, without any of the potential IP Address security issues.
Danny Asks: Is there anything we can do externally to help you bring back keywords? Can we ask Larry? Can we just click on more ads? (We all laugh.)
The conversation shifts to the back linking tools in Google Webmaster Tools
Greg Boser, Independent, Consultant (@GregBoser) complains that we (the online marketers) have to put in a lot of work to get a back linking dataset that is actionable from Webmaster Tools. He says the tool is lacking and really takes a lot of work with external tools to parse whether the links are passing PageRank or nofollow.
Duane Forrester, Sr. Product Manager, Bing (@DuaneForrester) says he has a solution to Google’s Not Provided keyword dilemma. He says to “Seriously. Use Bing more.” We all laugh, but then he continues, in all seriousness saying “We don’t have a problem sharing data.” He says that this “Not Provided” Google keyword deal started over a year ago; it’s not really new.. it’s been happening for a while.
Danny Sullivan notes that keyword data is not really “crucial data,” and that it’s just the fact that it’s data that we used to have that was taken away from us that is the issue.
The conversation is organically shifting around between keywords, and privacy, and competitive research and Webmaster Tools…
Rhea Drysdale, CEO, Outspoken Media, Inc. (@Rhea) chimes in and says: “You almost never have to do competitive analysis if you have enough data to do a deep dive into your own site.” IE: Your own referral keyword data; your own internal search queries; your pages that are getting engagement, etc.
Greg Boser says: “I am all about keeping the web secure” then continues by asking the crowd “how many people would move your site to HTTPS if you could get your web referral data?” (lots of people raise their hands). Greg says he thinks the Google “security” excuse is a cheap and an easy excuse and that the whole Web would be a much better place if everyone were encouraged to move to secure servers.
Greg Boser directly asks Brian White: “What users are being harmed from anonymous keyword data being passed??”
Brian says – briefly and kind of avoiding the question (for whatever reason; I’m not sure the reason): “There’s a lot written out there about what could happen. I’ll just say that.”
Danny Sullivan hypothesizes three reasons why anonymous keyword data could possibly theoretically be a security threat:
- Search info can be coupled with an IP address or other search data that could put your identity or safety at risk
- When searching on an unencrypted connection people can eavesdrop on a whole string of things you do.
- (the most unlikely of the three hypotheses): With Gmail and Google becoming more interconnected there’s a small possibility that a user could try to search through his or her own private email for something sensitive (like ‘Danny [insert prescription drug information + Social Security Number]’’ and, since Gmail and Google are more interconnected they may accidently enter their query into Google rather than Gmail. (He notes this option is very unlikely, but could be a hypothetical scenario.)
Greg Boser responds first by noting how cool retargeting is.
Changing the topic a bit organically Duane suggests: “Wouldn’t it be better if you could predict where people are going to be and what they’re going to be doing next? It makes more sense for us to be putting ourselves in front of the user”
Gear shift – the panel is now talking about the Hummingbird algorithm.
The Hummingbird Algorithm
Danny opens the conversation by defining Hummingbird, as he understands it. He says: Hummingbird is all about tying together entities and bringing entity search to the next level. It isn’t a “we don’t use links anymore” algorithm, it’s more of a “we’re still looking at all these signals, but we’re also looking at more streamlined, refined signals” algorithm update.
Brian White responds: First he switches between “framework” and “infrastructure” trying to decide which is a better way to describe what Hummingbird is. Then, deciding “infrastructure” works best, he says “it allows us [Google] to rank things a little differently and incorporate this notion of concepts.” For instance, you may do a search with a set of search terms and instead of scrutinizing those search terms and trying to find matches with those terms , the search engine/algorithm is working to focus more on the big picture concepts like what the user needs based on what they asked for.
He references an example. Say Danny enters into Google “buy an iPhone 5 near me.” Post Hummingbird, with this query the Search Engine says to itself “he doesn’t want to buy anything online, so maybe we should give him more brick and mortar solutions.”
Brian White references the Star Trek computer as a simple way to think about where Google wants to go. You say something to it, and it gives you exactly what you want now. It doesn’t give you a list of options; it helps you solve problems/queries immediately.
Brian mentions how amazing Google Now is (Note: Google Now has been a hot topic at this conference!)
Duane pipes in to say: It’s not a Google v. Bing world; often both engines have to be solving for the same problems, so they have to follow the same paths to get to the right solutions
Predictively – a way for search engines to better understand what gestures and conversational searches actually mean. This is where Duane thinks search is really going.
Duane says he sees the next generation of search like this: Someone has an iPad, they look at a picture of the Empire State Building, they tap the picture – the speakand ask “what can I drink there?”
It’s about making connections and understanding gestures, context and demand.
Understanding entities helps us crack that.
The idea is to help search engines get to the bottom of what you actually are asking for; what you want when you want it.
The panel speculates this might turn into “predatory aggregation to a whole new level.”
Greg Boser changes gears a bit and says he would change the name of the “Knowledge Graph” to the “Other People’s Knowledge Graph” if he could because the KG is really pulling in other people’s knowledge… knowledge from Wikipedia or Freebase (etc.).
“Google hasn’t created an original piece of content ever. That’s not what they do.”
Generally Google shows the source, but not for all sources and all queries all the time. (there’s some contention within the panel about this.)
Greg Boser and Rhea agree: If you need Organic Traffic build as many direct connections with your audience as possible so you can bypass Google.
Audience building is critical; building newsletters; distributing apps – all of that is critical. Organic is going away more and more. Search Engines are learning more and more and pushing to answer more and more of the questions themselves (EG: Knowledge Graph); They take your schema (entities) and use that information to inform their knowledge graph which can cannibalize your organic traffic.
Greg says the search engine that 100% answers your questions more than redirects is right around the corner.
There’s some discussion about Bing’s Snapshot pulling information in a way that is similar (but not the same as) the Knowledge Graph.
We’re circling back to the idea of Google as the Star Trek computer; the all-knowing fast answer machine that gives people the answers directly and doesn’t redirect people to secondary websites (RE: Your website).
Duane says: We’re going to see some pretty dramatic changes in the next couple years. For one, mobile growth rates. It’s predicted that mobile searches will very soon surpass desktop searches.
Complaining about a search engine that doesn’t work the same as it did five years ago doesn’t make any sense; instead, we need to be worrying about how people are changing — how technology is adapting — and we need to figure out how to change and adapt with it.
Greg circles back by saying – what I think is an INCREDIBLY IMPORTANT TAKEAWAY: You need to really focus, bottom line, on how humans seek knowledge. “I don’t care what form that takes or what it looks like,” it’s all about finding out how people seek information and ensuring that people get that information in front of them. It’s important to note that this doesn’t necessarily mean they use Google or keywords.
Link building now and in the Future
The problem is no one knows what a safe link is anymore. We’re all scared about when to nofollow and when not to, and how to disavow.
Brian White is talking about links in press releases now: The intention of press releases, historically, has been to get that information out so that people can read that report and then create content based on it. The press release isn’t necessarily supposed to be the content itself. It all leads to a bigger point – the quality guidelines Google has focus on linking with a purpose and making great content that people want to link to.
Greg Boser chimes in: Organic link building is really contingent on audience. If you have an audience you don’t have to go out and solicit links; you just build great content and people will link to you.
If you don’t naturally have a huge audience, you need to leverage the audiences of others to help build your audience. (How do you do this? Greg doesn’t really give recommendations but it’s an interesting conversation/strategy topic.)
RE: Disavowing Links
Greg says he wishes that there was a system in place where spam links just wouldn’t rank, and then by not ranking, spammers would learn their lesson and stop. Brian White says it’s not that simple, and that they have based their current system on lots of feedback
To finish up Danny asks each panelist to share their 30-second 2014 SEO takeaway. Here’s what they said:
Rhea said: We’re not search marketers or keyword chasers, we need to start thinking of ourselves as behaviorists. We need to change our search behavior as searchers are changing. I get bored, so I am glad my job changes, and is changing and I would encourage you to be glad, too.
Brian White gives a quick nod toward Mobile.
Greg Bose wants to kill the m-dot and leverage mobile to make/create more direct connections with consumers. He speaks quickly about whether you should go with a responsive design or an app and basically recommends both, as appropiate.
He also reminds us that “when the Star Trek computer takes over, he who has the largest connected audience survives.”
Duan’s final thought takeaway is all out thinking toward the future. In fact, he says that it’s incredibly important to think toward the future. He gives an example of a car company and says that they have an “in-house futurist” whose job is to forecast what kind of company they should be 10-20 years from now. He notes that in 10-20 years people aren’t going to be as excited about driving cars so the car company needs to adapt to the market (and create) cars that are lifestyle products. IE: Products that help people (specifically today’s millennials) in other ways besides transportation (he lifts up his phone).
Pay attention to millennials, he reiterates; the people who are coming next; those are the people who will really shape your business and determine if it survives.
0 Comments