|HTML tables versus CSS - worth the conversion?|
| 11:05 pm on Jul 23, 2009 (gmt 0)|
Hi I have a 6-7 year old site which is written using basic html & the odd table here and there.
Im thinking however of converting it all to css.
Apart from the obvious benefits of CSS (faster page load, put keywords at top of html etc) are there any obvious effects or advantages that a site built in CSS has over a table based site?
As mentioned my site is 6-7 years old and has ranked first page for its keywords for a number of years.
That said I cannot seem to break top 3 so am trying to gain any little bit of advantage I can get.
Is changing all your code to CSS risky? Would rankings drop?
I guess what I am asking is it worth the risk?
| 4:29 am on Jul 24, 2009 (gmt 0)|
The benefits I have found are directly related to a higher text versus code ratio, and better control over positioning of content on the page in code. Certainly div / css offers the opportunity to place text nice and high up in page code when planned properly, which can certainly help rankings to an extent.
How much that process would help you would be relative to the current code on the pages, and how well you can execute the changeover.
| 8:57 am on Jul 24, 2009 (gmt 0)|
thanks for that reply.
your 'higher text versus code ratio' hit the nail on the head for me.
I'd like to give google a nice feast of nicely written content at the top of the doc instead of a mess of content and table tags. although my site has ranked extremely well for the last six years. have to have a think to see if its worth it ;)
| 9:06 am on Jul 24, 2009 (gmt 0)|
This kind of conversion used to give a good ranking benefit. But as the years have gone by, Google has improved at not depending on the source order of the content, or on old-style text to content ratios.
There are many solid reasons that you may want to change your basic page layout away from tables. However, in my recent experience, you should not count on getting better rankings just because of making that change. In the process of making that change you may improve other factors about your page -- and that can help.
One of the problems with some table-based pages is that phrases and other bits of content that are visually related to each other get far separated in the source code, so that semantic relationship is obscured. In such cases, some added traffic may result from a changeover.
| 9:49 am on Jul 24, 2009 (gmt 0)|
One can redo the past... with minor benefits... or one can generate new (with CSS instead of tables) and get on with it.
As far as I can tell Google gives no credit or discredit to the coding on pages as long as the CONTENT has value.
I have a number of sites. Some are elderly tables, old time frames, fully css, and I don't see many differences in their listings.
| 5:20 pm on Jul 24, 2009 (gmt 0)|
" However, in my recent experience, you should not count on getting better rankings just because of making that change."
Based on what I see and control testing I have done, recently, I would have to disagree. Although it is not a prominent factor like selecting correct titles or the types of links you build, the difference, all other things equal between position 1 and 2 can and often is these types of factors.
| 1:24 am on Jul 25, 2009 (gmt 0)|
Interesting Cain - thanks. If I'm ranking #2 I'm hesitant to test anything at all ;) so I can't contradict you on that.
| 8:08 pm on Jul 25, 2009 (gmt 0)|
"If I'm ranking #2 I'm hesitant to test anything at all ;) so I can't contradict you on that. "
Nice point. With the way Google is fluxing most of the time it's not always priority for those with positions to 'rock the sweet spot' :)
| 9:18 pm on Jul 25, 2009 (gmt 0)|
I'm also pretty sure that Ted, if not others are by now using tables with CSS, as opposed to an either/or scenario?
CSS is not always about source ordering, it can be used to minimalise your table markup too, maybe that's a factor in finding the "sweet spot", maybe it's not.
I'm hesitant to recommend a complete immediate change due to may other Algo fluxing factors rather than anything against CSS layout conversions. It really is just about common sense, and why change what's working. But in order to keep up with fast changing situations perhaps it's better to think about which bits to changeover/migrate slowly.
Best advice is to think about more than a screen presentation, how would your table layout hold up then, which factors do SE's think are more important? - Imagine someone viewing your site on a small screen perhaps.. is the table causing a large horizontal scroll, if so is that working for your users? or would you/they rather the site was displayed vertically (regardless of source order).. what's easier for humans is usually like by the SE's too ;)
The above is almost a moot point anyway, as with CSS AND Tables you can make your table display however you like, vertically, different positions you name it, it can be done.. caveat only when IE6/7 are no longer around,
so meanwhile I think the decision is how can I use CSS to help my markup/users/SE's rather than a total "conversion" ;)
| 10:00 pm on Jul 25, 2009 (gmt 0)|
|I'm also pretty sure that Ted, if not others are by now using tables with CSS, as opposed to an either/or scenario? |
Absolutely, yes. I'm talking about putting down a layout framework with just one table, and then filling in the cells with divs and CSS.
Several years ago I went with full CSS positiong for a while, and the development time required to get a cross-browser success was just too much for me. I got tired of all the div content overlapping into other divs, backgrounds not filling out a div, etc.
I still use some of the successful all-CSS templates I worked up during that period, but for practical business reasons, I went back to a single table/grid for the basic layout and metally told the browser builders to "Call me when you've got yuour act together."
Some my table-based templates still use absolute positioning to place a content div at the top of the source code. And these sites, plus what clients bring me, are the reason I say that source-ordered content and text-to-code ratios aren't the big deal that they once were for Google rankings.
Google could care less about it - they are working very hard to deliver the best naturally relevant results to their users and they are getting pretty good at focusing on the content rather than the trimmings. The linking games are another story, but when it comes to on-page factors, CSS-P versus table cells is not really in the game these days. You can succeed with either approach, and you can screw up big time with either one, too.
This doesn't mean that I recommend deeply nested tables or garbage like that. I value semantic mark-up because it can signal the relationships between the different words and phrases on the page more clearly.
| 10:24 pm on Jul 25, 2009 (gmt 0)|
|Some my table-based templates still use absolute positioning to place a content div at the top of the source code. And these sites, plus what clients bring me, are the reason I say that source-ordered content and text-to-code ratios aren't the big deal that they once were for Google rankings. |
absolutely agree, but however I do feel you and I are caught in the "self fulfilling prophecy" part of that having naturally incorporated both, or bits of both as it suited us/clients as time went on.
e.g. I still advocate content first in source, though not with table tricks ;) the difference being in my case it's not and never has been about the SE's it's about the users and what I would like them to see first. With SE's it was a fad they thought would look good and it sounded nice that they were looking to users.
Now like you , there is no way on this earth I would recommend nested tables, instead use the proper HTML markup and corresponding CSS inside the table 'frame' if that what works best. That way means there's no table summaries for a browser se/reader to read through and e.g. form labels still associate with their controls etc.
We've come a long way from using tables for everything. SE's always have and always will have to keep up with emerging technology and are getting smarter too, but instead of the "what should we do to keep them happy" thinking, perhaps some should remember that's there is no "magic bullet" and it's due in no small part to hard work, common sense, and a good site as opposed to markup tricks.
in answer to the OP, Yes it's worth the conversion, but perhaps not all in one go? i.e. don't scare those SE's if that's the only goal! - even SE's are scared of change apparently :)
| 10:31 pm on Jul 25, 2009 (gmt 0)|
|it's about the users and what I would like them to see first |
Oh yes - so many compliments from users (and clients) about the way "the information just comes thundering down into the screen."
Even without those compliments, the site stats tell the story of low bounce rates and high pageviews-per-unique. It's not clear to what degree those factors may affect the algo directly, but happy visitors often mean more natural backlinks - and that's a good thing.