Gary Illyes (Google) and Duane Forrester (Bing)<\/figcaption><\/figure>\nGary: People are thinking about what Bing or Google wants. That\u2019s the wrong approach. Most sites and businesses that don\u2019t think about millennials will be going out of business. He\u2019ll add to Duane\u2019s description that millennials have short attention spans. If you can keep it to five minutes, then you\u2019re in a good span.<\/p>\n
Duane: There\u2019s also the amount of time you have to get their five-minute investment. That\u2019s seconds. Millennials will see the largest transfer of wealth when their parents die, in the next 10 years. This is a generation that was pounded by the recession, stayed in school longer, got more advanced degrees, and are getting higher paying jobs. This is the money you seek. Are you ready for a day when there isn\u2019t such thing as a search engine as we know it? Apps are like vertical search engines. Millennials are used to that search experience. Search is the data layer powering the experiences. So how are you working your way into that? How are you getting your work in front of these people in a meaningful way to drive value or save the existence of your company? Typing into a box is soon going to be a quaint idea.<\/p>\n
The topic transitions to voice search.<\/strong><\/p>\nYou can do voice search where the context flows from query to query. Apps like Google Now and Cortana are expected to answer these follow-on, context implied queries. The answers come from public data. If you\u2019re building your business around public data, that\u2019s not the best idea. Provide something more that users want.<\/p>\n
Any stats on if people are using the Cortana search button?<\/strong>
\nDuane: 100 percent of voice search happens vocally.<\/p>\nBest practices for content syndication?<\/strong>
\nFor other business purposes, it might have merit. If you\u2019re using it as an SEO tool, or link building exercise, that\u2019s not a good thing. For duplicate content, Bing isn\u2019t at 100 percent correct attribution. A lot of smart things can be done to signal the original content to search engines. Canonical. A block that says that the content is from a syndication partner.<\/p>\nGary: There\u2019s no duplicate content penalty. If many pages duplicate the same content, they pick the canonical from the cluster of pages. A bigger problem is if you think you\u2019re doing something legitimate, like selling your content to another site, and you might have problems.<\/p>\n
What about HTTPS for ecommerce sites?<\/strong>
\nGary: Don\u2019t do it for the ranking boost. Do it for users. You never know where your user is.<\/p>\nDuane: It\u2019s tough for engines to take a line on security. We\u2019re all in for safety for our own users, employees and customers. If you want to wrap your head around these concepts, read a book called \u201cFuture Crimes\u201d by Mark Goodman. This ex-FBI agent explains how porous everything is and how far and fast data spreads. This explains why search engines are recommending it. This book keeps Duane up at night and made him depressed. Private browsing in a Firefox window \u2014 anything like this is window dressing to security. That\u2019s the level of problem solving they\u2019re dealing with at the engines, and seeing how they can rationally influence what\u2019s happening to businesses.<\/p>\n
Will Google Search Console and Bing Webmaster Tools treat HTTP and HTTPS versions of my site as being part of the same domain?<\/strong>
\nGary\u2019s evasive.
\nDuane: He pushes for things like that. Tools are powered by a series of other services that the search engine uses. The net effect is that they may want to integrate that, but it requires other teams to make changes.<\/p>\nWhy is a scraper site outranking me?<\/strong>
\nGary: When a scraper copies your site, search engines are pretty good at discovering the original source. If your site is not ranking for your content, there might be something on your site dampening its rankings.
\nDuane: He gives a warning about using DMCA requests too much.<\/p>\nWhat’s the best URL structure or site architecture for an international, multi-language hub site?<\/strong>
\nGary: It doesn\u2019t matter. Use locale and language identifiers. It doesn\u2019t matter in the end. If you use hreflang, you can specify exact language and locale. It\u2019s not necessarily good to think about what will be good for a search engine. Think about what users will want.<\/p>\nDuane: It depends on your business. If you have a big office in France, you probably want a .fr. If you\u2019re a U.S. business who wants to touch an area, then you may want it in a folder. It\u2019s all about giving the search engines as many signals as possible. The folder may be the way to go so that all the content is collected for you. This page is for this product for customers in France, please rank it as such. This is a level of audience segmentation that you need to know about. In France, the mindset is very nationalistic. They default to it as much in their daily life as possible. So, you may find that sites with a .fr would be more trusted by a French national.<\/p>\n
How is page speed playing into Google rankings? The PageSpeed Insights tool only ever recommends browser caching.<\/strong>
\nGary: Page speed is great. Optimize your site for your users because users like fast sites. We also talk about attention spans, because people demand the content fast. How do you create fast sites? Web Speed Test, a bunch of resources you can check for making a site faster. You can compress JS, CSS, HTML, convert images \u2026<\/p>\nDuane: (He picks up where Gary leaves off.) \u2026 Less dependencies on third parties. Here\u2019s one more way speed matters for SEO. The faster a user clicks back from a SERP click-through, the faster you\u2019ll rank lower.<\/p>\n
Danny picks up that this is an admission that Bing uses click data for ranking.<\/strong><\/p>\nDuane uses a suspicious voice and says\u00a0\u201cDo we?\u201d and he says he doesn\u2019t disagree.<\/p>\n
Gary: Google uses click-through data when they\u2019re experimenting with new features. Google uses click-through data in personalization and disambiguation (like searches for Apple). Other use of click-through data and user signals is a real pain.<\/p>\n
Duane: If you look at all of the data, that\u2019s a lot of noise. It\u2019s not easy to use user signals. You can\u2019t trust the data you\u2019re seeing because there are outside influences. Bing also looks at click-through data in experiments to see how it influences usability.<\/p>\n
Why does the number of results on a Bing SERP change?<\/strong>
\nIt\u2019s experimentation. The number of items on a page generally differs by topic. There\u2019s an above-the-fold page real estate issue that they have to deal with. The more they understand data from a source, the more comfortable they feel answering the query.<\/p>\nReferrer spam in analytics \u2014 when will Google handle that?<\/strong>
\nGary: The issue is being escalated. There\u2019s a plug-in you can use now.<\/p>\nWhy did Google go from 7 to 3 in the Local Pack?<\/strong>
\nNo answers.<\/p>\nPenguin and Panda?<\/strong>
\nGary: Panda is rolling out. Gary checked it yesterday. The new Panda update will be a multi-month update. (Note: This is news!<\/em>)
\nRegarding Penguin, Gary says it\u2019s rolling out in the foreseeable future. He won\u2019t give a specific date or else Barry will get him in trouble again. The next Penguin will be a real-time algorithm factor, and he is hoping it will be before the end of the year. He\u2019s not working on Penguin, though.<\/p>\nSafe to use anchor links on page?<\/strong>
\nYes.<\/p>\nShould we avoid giving you too much structured data because you\u2019ll cut our pages out of SERPs?<\/strong>
\nGary: The assumption that a search engine will drive you out of results for rich data is wrong. Sites getting featured snippets see increased clicks.<\/p>\nDoes using incognito search give impersonalized SERPs?<\/strong>
\nEverything is personalized at least by geographic region.<\/p>\n