At SMX West, currently underway in Santa Clara, there are a number of sessions involving blended search. Two of the most obvious blended search results are images and videos, making them harder to ignore. Google took away one huge click that people had to make, bringing images and videos out into the main search results. They’re still relatively rare: you can make quite a few queries without discovering images and videos. Speakers Todd Friesen, Eric Enge, Benu Aggarwal, and Henry Hall brought up quite a few reasons why you should improve your image and video optimisation for the sake of them appearing in regular SERPs.
One thing I find particularly interesting about blended results is how each feature is treated differently, either as an additive or a subtractive result. Videos are subtractive, whereas news stories are additive. However infuriating it may be to find yourself removed from an optimal SERP position due to a video, it isn’t the case that you can’t do anything about it. At the least, if you see a video ranking for your keyword, you know that Google gladly includes blended results for that query. Given this knowledge, you can attempt to rank your own blended content.
One great piece of information from this panel was the change in how people look at search results pages when blended content is included. These heat maps show how much more interest people seem to take in the results as a whole when they are presented with a range of content. This is great news for multiple people: Google can be pleased that people are finding interesting content in all of its results, as they’ll likely keep coming back if they feel they’ve been given a whole page of good resources. It’s good news for SEOs who have traditionally believed that no one looks beyond the first three results. It’s also good for all of us – search engine users – who are apparently finding the whole page useful.
While no one could give us a sure-fire way to have images appear at the top of a blended results page (and how could they; if we all knew how to do it, there would still be limited real estate at the top of the page), they brought up some interesting points about the oddities of Google’s image listings. For example, the images that Google returns as blended results aren’t the top results from the images’ page. They also sometimes appear to be a bit “random” and not necessarily optimal. Take a look at a search for “pictures of santa clara“: if you receive the same results as I do, you’ll see a rudimentary map, a slightly better map of the entire county, and a picture of the mission of Santa Clara. I don’t agree that any of those results are particularly helpful… if I’d made that query, I’d probably want pictures of the city. Also compare these to the regular image results. Only the mission picture appears on that page.
To me, these less than optimal results within blended search suggest that Google has a little way to go before the images it chooses are totally relevant. While determining relevancy and quality must be harder than with written content, I have every faith that Google can sort it out. Of course, three pictures can never serve everyone’s needs, but no one is going to be satisfied with the Santa Clara blended images.
I would have thought it went without saying that general optimisation of a site as a whole goes a long way to having images and videos show up in blended results and image searches. Perhaps it needed to be repeated, because there were quite a few mentions of basic concepts such as including good meta information, well-worded alt attributes, titles and URLs. Really, there’s no problem in emphasising what we should all be doing already, because we don’t know where one additional piece of information will make the difference between a great listing and invisibility.
