Showing posts with label real-time search. Show all posts
Showing posts with label real-time search. Show all posts

Monday, December 14, 2009

Google gets Personal, Quick, and Snappy

...but have they also got sloppy? Last week, Google made three interesting announcements regarding search marketing. Here's a quick summary, and some thoughts around these:

1. Personalized Search
It seemed inevitable that search needed to evolve from a general to a more personalized level. Google's Personalized Search attempts to do just that by using past search history to serve up more relevant results for each individual. This is the case regardless of whether you are signed into Google or not, and it seems that Google have made the opt-out process quite difficult to figure out (not to mention that most searchers will likely be oblivious to it). This, of course, brings up the question of privacy, even more so since there has been talk of cookies being opt-in in Europe in 2011.

What does this mean for SEOs?
Results can now rank higher or lower based upon an individual's past click behaviour rather than just the search algorithm. This makes SEO harder to optimize for (even more so than before) and can make SEO measurement more complex; measures of success will need to move beyond a traditional rank metric.

To see more, watch Google's Personalized Search Video.

2. Real-Time Search
Relevance has always been a big factor in good search results, but with the continued rise of social media, recency is becoming more and more important.

Google's Real-Time Search results sit in the middle of the SERP and automatically flow as new information comes in, mainly from Twitter (though it also includes results from Google News, Google Blog Search, Friend Feed updates and others). Alternatively, you can also click on "Show options" and "Latest" to see a page just with real-time results. Here's what it looks like:

Example of Google Real-Time Search Results (Show Options > Latest)

Although the results are integrated into the SERP, execution doesn't seem to be the cleanest - which begs the question of whether Google launched real-time search as a knee jerk reaction to Bing. Many reviews I've read have also commented that the results are below par and can be easily influenced, though Google claim that only information that is highly relevant (and not spammy) are included... though actual results suggest otherwise.

Other questions also arise regarding Google's current implementation. For example, with the real-time search results appearing slap-bam in the middle of the SERP, will this mean more people investing in PPC in order to gain more control of their presence above the fold? And will it encourage people to spam more to try and influence real-time results?

What are the implications for search marketers?
Regardless of current execution (which will hopefully improve), search and social are becoming more and more intertwined. Quite simply, web marketers need to get their search marketing and social media strategies in order as the two continue to converge.

Read more about Real-Time Search.

3. Google Goggles
Google Goggles (launched in Google Labs) allows you to simply take a picture (with an Andorid phone) to search the web. It doesn't work for everything and a valuable business application right now is questionable. However, it is pretty neat in a geeky, experimental kind of way.

Some of the things you can do include:
  • Finding out what businesses are nearby
  • Identifying objects, places and landmarks
  • Identifying products through barcodes
See Google Goggles in action.

Monday, November 2, 2009

Search Goes Social

Last week was an eventful week for search and social marketing. First there was the announcement that Twitter had signed deals with both Bing and Google to give them access to its full feed of public tweets. How exactly the search engines intend to use this real-time information from Twitter remains a little fuzzy, though Bing does already utilize some of this in Bing Tweets.

On the same day Google announced a separate "Social Search" experiment in Google Labs. Google's Social Search displays links and updates based upon the searcher's own social network. The idea is to show personally relevant search results based upon your social connections. Why? Because people trust information more from people they know or are connected to.

The results (currently appearing at the bottom of the standard search results page) are based on various information sources, such as:
  • Gmail contacts
  • Google reader subscriptions
  • Social networking profiles on your Google profile
  • Friends on other services, including Flickr, FriendFeed, Digg, YouTube...
These search results also tell you how you are connected to that person; a nice touch.

There has been much talk about the growing convergence of search and social as well as predictions of their marriage in the future. But these are promising signs of the major engines taking serious steps forward where social media is concerned.

Learn More

To check out Google Social Search yourself, go to Google Labs and join the Social Search Experiment.

To understand How Google Social Search Works, watch the YouTube video by Matt Cutts.

Monday, September 21, 2009

Real-time Search Wars: Google vs. Bing

In a previous post, Why Should Search Care About Social, I talked about search engines needing to better incorporate social and real-time search results as part of their Search Engine Results Pages (SERPs).

Since then, there have been a few developments by both Google and Bing. Here is a very preliminary review of progress so far:

Google
Once the SERP shows in Google, you can expand out a "Show More Options" link to provide you with time-specific results by "Recent Results", "Past 24 Hours", Past Week", "Past Year", or a "Specific Date".

However, "Recent Results" are not very recent, at least for the real-time searcher's needs. Searches I've tested can vary from 3 hours, to 9 hours to 3 days ago, which is a lifetime in the social world (maybe this is the reason why they keep it relatively hidden).

Search Engine Land recently revealed a Hidden Google Feature to find out what's new in the last minute or second. However, there are still kinks to work out, as revealed the day after the VMA awards:

@adamvonwillis RT @dannysullivan: still how "real time" is google's search w only 4 matches on kanye west in past min - How pathetic!

We need to cut Google some slack though. Since it's a hidden feature, it will be work in progress, and I'm sure this will be much improved once it is officially revealed to the public at large.

Bing
Bing have a Beta site called Bing Tweets which claim to "Fuse Twitter Trends with Bing Insights".

Bing Tweets provide a way to learn more about what trending topics are about on Twitter, by showing Twitter Trend topics side-by-side with Bing search results on those topics. It's a nice step forward, but mostly comes across as separate Twitter-specific and Bing-specific elements, lumped together on one page. Here's a breakdown of each element:

Bing Tweets Screenshot

Twitter-Specific
The top left of the page displays the trending topics on Twitter, along with an RSS feed showing tweets mentioning a highlighted trending topic down the left side. A nice touch is being able to view topics by either "Popular Now" (all topics), "People", "Places" and "Products".

Bing-Specific
The majority of the Bing Tweets page is taken up with the standard Bing Search Results, using the highlighted trending topic as the keyword. The results, or course, include universal results such as video.

Search Bing and Twitter
The "Search Bing and Twitter" function on the top right of the page allows you to conduct your own keyword search so that you are not limited to just the top trending topics. This is nice as you can do searches on your own company and products, for example.

Share This Result
Bing pre-populates a message within "Share This Result", leading people back to the Bing results page. They also cleverly utilize a #bing hashtag, helping to increase their Twitter exposure.

Probably the nicest part about the "Share this Result" feature is allowing you to easily share the page via Twitter or via other popular social media vehicles (including Digg, StumbleUpon, LinkedIn, and others).


So overall, some nice, positive steps forward, but there is still a long way to go to winning the real-time search war.

Thursday, September 10, 2009

Why Should Search Care About Social?

As Search Marketers, we have an amazing amount of insight into metrics and are able to link every dollar spent back to ROI, or to a Key Performance Indicator (KPI). The same can not be said about social media.

So why, as Search Marketers, should we care about social media?

1. Link Building
The obvious reason to care about social media is for link bait to your site. A few examples may include getting people to Digg or StumbleUpon your content, sharing links to your site via Twitter, or through your Facebook fan page. However, the link between search and social goes much deeper than link juice.

2. Customer Support
Product support-related searches on Google will often include Forum results - people discussing you, your company, your products, your customer care. A search on "ipod forum" brings up 139 million results. Whilst "pc help" brings up 240 million. These may be extreme examples, but wouldn't it be nice for your own company-hosted forum to appear in the top rank when someone typed in a support-related query for your company or products?

3. Social Sites are Search Engines
Google may be Number 1 when it comes to search engines, but YouTube is the second largest search engine in the world. Additionally, immediate events (real-time search results) aren't easily searchable outside of Twitter.

Web users aren't stupid. They are getting more sophisticated and adapting their search behavior outside of the traditional engines. For example, if I want to know what conversations are happening at this exact second, then I search on Twitter. If I want to look for a good deal on trinkets, then I search on eBay. If I am looking to rent an apartment, then I search on Craigslist.

If search is expanding out to all these different platforms, then we need to think and work outside of just the traditional search engines. It also means we need to optimize more stuff so that we show up in these searches. For example, making sure we include keywords in tweets, carefully crafting video titles, and using metadata around video transcripts... the list goes on.

4. Universal Search to Include More Social Media Results
Engines recognize that people are getting street smart when it comes to search, and they aren't just going to sit by whilst searchers wander elsewhere.

ComScore found that universal search results are increasingly dominating search engine results pages (SERPs). Universal search (also known as blended search) shows you video, images, news, and shopping, specifically broken out as part of your search result.

Last month, Matt Cutts told the SES San Jose audience that Google sees reviews as an extension of search. So could a future addition to universal search results include product ratings and reviews? And what about Twitter feeds, Facebook status updates, and FriendFeeds? As search marketers, we need to stay on top of this if want to continue capturing shelf space in SERPs.

5. More Forward Thinking: The Future of Social Search
There's no question that Google is great at a lot of things. But right now, it doesn't do such a good job of making sense of social sites. Looking forward, how will this develop and how will search and social become better integrated?

Groundswell author, Charlene Li gave an eye-opening talk on how to prepare for the future of search, at SES. In her presentation she talked about search engines relying much more upon social networks to deliver results that are a lot more customized to each individual searcher.

Some key points from the discussion included:

  • Better real-time search results
  • Personalized search results based upon social data
  • Searcher intent becoming clearer with geographic location, time and social context
  • Search recommendations based upon reviews from people we know, rather than from complete strangers

Just think - you're stuck in your hotel room in a strange city, and your stomach is growling. You need to find a place to eat. The search engine detects your geographical location, and recommends a restaurant two blocks away, based upon positive reviews from friends within your social network. Heaven!


These are just some of the reasons why search should care about social. I am sure there are more (feel free to share ideas). It also serves as a good reminder that moving forward, people need to be at the center of our search strategy, not keywords.