Matt Cutts gave an in-depth SEO Q&A session at SES San Francisco.

Q&A with Matt Cutts at SES San Francisco: Social rep for SEO, traffic cannibalization& Google transparency

to read

Editor’s note: This article initially featured a Matt Cutts misquote about Google’s ability to crawl Twitter, which has been updated as of August 16.

The most famous man in search delivered the opening keynote for SES San Francisco day two. Matt Cutts offered insights on Google’s shift toward the Knowledge Graph, social’s role in SEO and transparency for webmasters. He also fielded some tough questions from marketers on the rise of answers IN SERPs and how this could steal traffic from sites. 

One of the key focuses for Google is moving away from being a search engine and toward becoming a knowledge engine. Cutts emphasized the Knowledge Graph has an increasingly important role in Google search (and some marketers were not pleased!) Check out Brafton’s coverage of the full Q&A session:

Matt Cutts answers a question at SES San Francisco.Question 1 (from an SES San Francisco attendee): Is there a plan to improve the technology to focus less on content to understand a company’s quality in terms of products and services?
Answer (from Matt Cutts):

Social will become an increasingly important way to get an understanding of what is the best quality. Links will always be part of that for Google, but Panda and Penguin will get more refined. Marketers and webmasters ask, “When’s the next Penguin update?” and maybe they shouldn’t seem excited. The Penguin updates will cause some SERP flux.

Q2: [Generally, the Q asked which social signals Google uses.]

Editor’s note: The initial release of this article misquoted Matt about Google’s ability to crawl Twitter. Matt Cutts has clarified that Twitter blocked Google from crawling for 1.5 months when the deal ended, and we’ve updated the article to reflect his answer.

A: Google can only use social signals from Open Graph sources. It can’t crawl Facebook pages to see who is reputable, reliable or has real world impact as a brand has on those platforms. 

Q3: What happened with Twitter? [Google and Twitter ended their relationship in 2011]

People were upset when Realtime results went away! But that platform is a private service. If Twitter wants to suspend someone’s service they can. Google was able to crawl Twitter until its deal ended, and Google was blocked from crawling those pages for 1.5 months. As such, Google is cautious about using that as a signal – Twitter can shut it off at any time.

“Google is cautious about using [Twitter or other private social networks it is able to crawl] as a signal – Twitter can shut it off at any time.”

We’re always going to be looking for ways to identify who is valuable in the real world. We want to return quality results that have real world reputability and quality factors are key – Google indexes 20 billion pages per day.

Q4: What’s with the link messages people have been receiving? 

Google is focusing on transparency. There was a Tweet about the link messages that said “it’s cheaper to go white hat in the long term” and that’s what Google is striving for. And by becoming more transparent Google can show it’s not giving special advantages to certain web properties. The link updates work toward that.

Notably, Google will not be not be giving specific info about its algorithms – that won’t become too transparent.

Q5: What about when you [Matt Cutts] said not to overly optimize?

A; If you’re doing good SEO, you’re focusing on content for users, you’re making your sites faster. My hold back on SEO referred to not buying backlinks. That Tweet about it being cheaper to go white hat is [essentially what Cutts aimed for when he said not to overly optimize]. SEO is good!

Q6: Is Google becoming less of a search engine and more of a publisher? Knowledge Graph provides information but is eating up traffic to other sites (especially Wikipedia) and Google Flights takes traffic away from Orbitz, etc. How many publishers are impacted? Should verticals be scared if they get on Google’s radar that Google will cannabolize traffic?

How many publishers are impacted [by Knowledge Graph and answers in SERPs]? Should verticals be scared if they get on Google’s radar that Google will cannabolize traffic?

A: Google is about answers. Wikipedia is focused on knowledge and probably approves. But for marketers that want to maximize traffic, the key will be to focus on content that has original value – such as original reviews, unique data or insights.

When a user types “What’s the weather?” there’s a straightforward answer Google can give. These are objective facts. When you want to get insights or reviews, such as “Where’s the best place to buy a car?” Google won’t be able to give a simple answer. Marketers have to focus on offering unique value for objective queries.

Q7: Where’s the line for what Google considers objective?

A: Users expect more every year. Users want to be able to find more information directly on the SEPRs. [Cutts had more to say on this. Typing fail!]

Q8: Google is doing right for itself, but how can others pay content writers? What about webmasters?

“For marketers that want to maximize traffic [in the face of Knowledge Graph and answers on SERPs], the key will be to focus on content that has original value – such as original reviews, unique data or insights.”

A: Google is focused on what’s good for the user. The search quality/ knowledge team doesn’t care about how much money Google makes – they focus on what’s good for users.

[The crowd is dissatisfied, Danny Sullivan offers a follow up question/comment.]

Q9: [from Sullivan] For sites that are built on sharing facts, Google IS taking their business. It may be good for the user, but if Google is extracting the content to do what’s good for users … At the same time, Google still delvers the majority of traffic to sites.

A: Google’s focus is on the best user experience, but we are mindful of webmasters. Users HAVE to come first or searchers will go elswhere, but we understand the web IS websites, and it needs to be good for webmasters. Otherwise people would move to apps and search wouldn’t work out.

In Google’s “10 things we know to be true,” we profess our dedication to focusing on the users. 

Q10: Is it possible Google would release statistics about percentage of clicks to going to organic results over Google products or paid results, as well as if its top 5,000 domains or little guys?

A: I can make no promises, even if that would be great. A way to do that is to accumulate the rankings and see what Sistrix calls the sites that move up or down as an imperfect indicator. I’ll pass that suggestion on because we are moving to transparency. [He suggests it will be difficult because of the sheer volume of queries Google handles.] We field 3 billion searches each month.

“Users HAVE to come first or searchers will ditch Google, but we understand the web IS websites, and it needs to be good for webmasters.”

Q11: [The question centered on how rankings have changed since Panda.]

A: There’s a site that measures how many times Google ranks original content as top results, and with Panda they are striving to achieve this.

Q12: Can you talk about the importance of Google+ since Google wants to be more social but doesn’t have access to other platforms?

A: We haven’t put a LOT of weight on +1s yet. Google wants to consider social data, but +1s aren’t everything – Google considers feedback about Search Plus Your World

Q13: Do Google products (ie: YouTube) give higher rankings? Should marketers publish on Google properties?

A: We want people to be able to trust Google [not favor Google in results unmerited]. 

Danny Sullivan joins Matt Cutts on stage for a keynote discussion as part of SES San Francisco on Wednesday.
Q14: From a content writer’s perspective, what are the key technologies in Knowledge Graph and how can we make it an SEO tecnique?

A: The primary part of what we’re building Knowledge Graph off of is Freebase. It remains open source. You can download the data from Freebase and use it yourself. If you see something inaccurate, you can report it. Google doesn’t fix that, but Wikipedia, for example, will. If you want to know the data Google has under the hood – on people, versus locations, etc. – you can check it out there.

Q15: Why doesn’t Google have a rating system? If you’re moving to transparency, why isn’t there a tool that shows what you’re doing wrong so a site can do better in its space?

A: Giving actionable data on how to make a site better is exactly the direction that Google wants to be moving in. If there’s spam, we want to say, “Yes, we found spam.” This speaks to the unnatural links warnings. Now we’ve started saying: we think the site is good, but we don’t trust some of these links. Hopefully marketers can see what went wrong so they can change it. That’s the vision.

Of course, sometimes the message gets confused [people panicked with the new link messages!] so the delivery needs to be good.

“Giving actionable data on how to make a site better is exactly the direction that Google wants to be moving in.”

It’s useless if we tell webmasters, “We don’t trust you,” but we don’t give insights. By the end of the year, we hope to offer more on that.

Q16: The SEO community is in panic since Panda and Penguin, and people are rewriting all types of content. Can you speak to this?

A: Panda went after low-quality content. It was a big change, but it’s what we felt was best for users. Last year, everybody was asking about Panda, and now the market seems to have shifted. People understand good content, research, unique content with value is essential to SEO.

It was then that we were able to look at linking, so Penguin came out. People were shaken by that, and sites that have social sharing, natural sharing – instead of buying links – are generally not going to be hit. By next year, we hope marketers/ webmasters get that and natural link building becomes the norm.

[To the Panda query about “rewriting all types of content] If it’s the same content on different pages of your site, repeating your own content yourself, you might not be impacted but don’t use the same boilerplate content on every single page.  

[Stay tuned for more SES San Francisco updates.]

Katherine Griwert
Katherine Griwert is Brafton's Marketing Director. She's practiced content marketing, SEO and social marketing for over five years, and her enthusiasm for new media has even deeper roots. Katherine holds a degree in American Studies from Boston College, and her writing is featured in a number of web publications.


  • daveintheuk

    Strange, they can scrape social connections though. Creepy.

  • daveintheuk

    The sad thing about Google rushing to give “quick answers” is that users miss out on discovery great sites, great content and more about the subject they were searching for… but then perhaps Google is smart, realises this and wants to keep users hooked on searching for “quick answers”. Worse still, in doing so it is taking huge chunks out of publishers incomes and reducing the amount of content producers on the internet. In some cases the simpler queries were publishers bread-and-butter earnings that allowed them to build the really good, deep content – by cutting off this income, Google is killing swathes of publishers off.

    • Evan Jacobs

      It’s tricky, because it’s in Google’s best interest for users to find the information they want as fast as possible… in part because the more time they have to perform unique searches, the more PPC impressions can be shown.

      I agree with you though, the emphasis on fast results means that good content can be buried and remain unseen in SERPs. However, publishers can alleviate this somewhat with a good social media strategy, as the multiplier effect of sharing will get their content in front of the right people.

      • daveintheuk

        It wouldn’t be so bad for publishers if Google didn’t give “their” content such massive prominence as it does with products like Knowledge/+ Local…

        I agree on social, and it works well for us and other publishers we know – but the fact is Google controls 90% of the search market here and that is how most people discover content. We can build our brand, built ultimately if somebody carries out a local search and everything above the fold is a Google property or the “official” site us independent publishers find it really hard and gets worse with every Google “update” in that arena. Sure if users *see* our name they will often click it – but we’re burred below the fold and above the fold Google is providing something that looks like the answer to the user’s search.

        Perhaps part of the problem is that Google pushes their product as “the answer” when in fact they are simply “another answer” – for example, how many Google+ Local pages are blank or very, very thin yet have massively preferential treatment over, better pages.

        The fact is Google knows their content is weak and automated, so rather than letting it rank alongside everyone else’s and subject to the same rules (how would “knowledge graph” content scraped from Wikipedia fare if subject to Panda!) they simple give it massive exposure and present it as “the answer”.

  • Peter Morris

    Makes me laugh how thy call it THEIR knowledge graph. Every SERP I’ve seen with a knowledge graph entity showing is just stuff they’ve scraped from Wikipedia! I thought scraping and duplicate content were an evil sin Mr Cutts. Oh sorry, not if you’re the biggest scraper of other peoples content on the planet I guess. I’m not usually a Google-basher, but that is total BS – they say users don’t want to click through to Wikipedia to see that content? Tough, create your own content then Google…

  • Brad Dalton

    Does this mean a lower time on site is better with lower page views per visitor?

    • Katherine_Griwert

      Brad that’s such an interesting question! I think it merits reflection. In the end, it seems so arbitrary to gauge those metrics according to how they help SEO – in a more big picture sense, interaction metrics tell you about different pages’ values with respect to how you want people to be using a site. Lower time on site would be expected for say, a glossary term page – and that’s a case where it’s a sign the visitor got what they needed quickly. But I’d expect a visitor to spend a longer amount of time on an in-depth blog post or reading a case study. And more page views could be the sign of discerning visitors reading all about a brand before making decisions about it …

      Google is trying to identify the sites that offer the best user experience, and it seems sweeping to make a statement about how time on site or average page views reflects UX. It depends on the intention of the site and the user, no? I know this didn’t really “answer” your question – but thanks for keeping the discussion going.