An interesting argument appeared on the Jitbit blog this week. In it, the author alleges the default "above the fold" area of a Google search results page (SERP) is only 18 percent “actual search results”.
This was, in fact, an interesting claim. The author's "reasoning" suggests on his 1280x960 resolution screen, the search results take up a box 535 pixels wide by 425 pixels tall, 18.5 percent of his window if you multiply the resolutions and consider square pixels to be the same as measuring something in square inches.
Yet it didn't stop there. According to the author, those paid search (PPC) ads aren't useful to him and neither are the rest of Google's properties, so they don't count.
The author continued to suggest that the page had "about 45" different links, highlighing the black navigation links, the "Sign in" button, all the UI elements, ad links and the search results. That's when his math gets fuzzy.
The author jumps around between different numbers of links, trying to make a point that out of all the links, only five were actual "search" results, leading to a claim that only 11 percent of the links are actual search results, then later suggesting an ads to results ratio of 8:7, "which is 47% of the links are actual results."
I consider myself technical, but still stuggle how an 8:7 ratio is anything but 114 percent or how the author went from 11 percent to 47 percent.
Yesterday, the author submitted the post to Hacker News, prompting a response from Matt Cutts late last night.
Google's Distinguished Engineer outlined three bulleted points on his "major issues with this article":
1) the left-hand column is counted as non-search, when the left-hand column is entirely about search. The left-hand column gives you ways to refine your search: you can limit the types of search results like news/images, slice/dice search results by date, limit search results to verbatim matches or to change the geographic weighting of search results, etc.
2) the actual search box is counted as non-search, as are the estimated results count and the time the search took.
3) the article treated whitespace as non-search, when shorter columns can actually make it easier and faster for users to scan the results.
Cutts went on to conjecture that Google thinks the ads "can be as helpful as the search results in some cases," defending the position with a Google statement dating as far back as 2004.
A spirited debate rages on, with more than 200 comments posted. Most of the debate centers around the semantics of "results" versus "search."
Whether you agree with the blog post or not, the author attempts to back up his claim by stating, "we're all technical people" as a disclaimer to all the geeky details. If we really are technical, let's start talking about view port sizes and fixed width web design.
In all seriousness, if you're going to claim percentage sizes based on pixels, at least discuss it based on viewport size, not the entire screen resolution. The view port is the actual canvas area where HTML can be drawn in a browser. Its size takes the browser's tools and scrollbars out of the usable screen real-estate count.
If we take the design into play, we see that the main content area is limited to 1250 pixels, with a center column content area (center_col div) as 556 pixels. By my count, that's already 44.5 percent. I won't go to far into the user experience gained through responsive design or the user benefits of utilizing white space in user interface design to offset various areas of a page.
Here's the same search on a 7-inch tablet.
This image certainly seems to be more about search than the ads or the links to other Google properties.
Let's not even delve into the fuzzy counting of links.
Cutts’ points remain true, as do the countless others who suggested there is a trend that Google's interface is less and less about organic search results. Google's position is they are doing what they think are best for their users.
Google's revenue shows paid ads are working. People wouldn't continue to purchase them if Google users didn't find them useful, even if it means cutting the organic resutls from 10 to 7. (Interestingly, that was done for about 18 percent of search queries. Hmmm...)
What do you think? Has Jitbit found its 15 minutes, or is the author onto something?