What Do Search Engines Really See?

A question has come up over at High Rankings Forum about where text should be placed on the page, or if it even matters. Many in the SEO community disagree on this issue. Some will say the search engines will find the text no matter what so it doesn’t matter where it’s placed, while others argue that putting text higher up in the code or “above the fold” give it more relevance.

Which is true, or are they both?

I believe the first is certainly true; the search engines certainly will find the text and index it and therefore weight it accordingly. But will text higher up in the code give it a boost? I’m sceptical, and responded as such in the forum.

The search engines see code so they only know where your text lies within the whole of the page’s HTML. They cannot determine where your text lies on the visible page.

While I believe my response is correct, it was a bit too simplistic and I did not provide enough information to substantiate my thinking. I’ll do so here.

Dan Thies, whom I respect very much responded to my comment, pointing me to two resources which discuss visual page segmentation. These documents show how the search engines may try to analyze the code of a page and therefore determine where those elements will display on the page and potentially how important each of those elements are.

I think these are very important documents to read for any serious SEO and to be honest I had not read these until Dan pointed them out. But I did know their concepts of page segmentation. I’ll let these documents stand on their own as their own authority, but I do want to address the nature of my comments and why I believe I’m still essentially correct.

I believe that visual page segmentation largely relies on traditional HTML for attempting to determine where each distinct elements of each page falls when viewed in a browser. Even if engines are analyzing CSS as some believe to be true, this will only give a partial picture, and without the search engines actually pulling open each page in a browser page segmentation will rarely provide a 100% true representation.

Let me use my own site as an example. My home page has four distinct elements, the header, the side nav, the footer and the main body area.

All but the main body area are consistent we we’ll assume that the body area will be given the most weight based on that alone. But where does the body fall when analyzing the code? At the very top left.

We’ve used CSS and includes so that the top, side and bottom areas are all pulled from a single file. These can still be segmented but if the code was being analyzed at face value then the search engines would tend to think that this information all comes last therefore making any analysis based on where the information displays on the page virtually useless because the search engine’s best guess is wrong.

I’ll throw one more thing into the ring. If my body area had a very large image that pushes the text down 500 or so pixels, this again will give the search engines a poor indication of where the content actually is viewable on the page.

This is still a pretty simplistic argument I know, so let me be the first to admit that I’m not 1/8th of the researcher that Dan Thies is. I’m open for discussion and to be proven wrong.