Forum Moderators: open
Would appreciate if anyone could break this down per SE
H tags certainly play a part in algo's of the majority of SE's and so must be a part of your armoury.
My best advice is to find pages that rank well on mid-range competitive keywords/phrases, yet are obviously unoptimized. Analyse these types of pages.
Take a look at this page [cpsc.gov] which ranks No 1 at Google for window blinds.
Good luck.
Edited by: NFFC
I've never heard anything about more than one H tag on a page either helping or hurting. I would also love some input on this, since I like to define the look of H tags in a css file and then go with the easy, consistent page formatting that permits. I've never seen evidence of a problem, but evidence can be lost in the "noise".
What I really wanted to know was if anyone had experience of what happens when you use multiple, say h4 headings, stuffed with your keywords. And whether you can repeat keywords amongst the different headings as you can in the meta tags keywords area.
A few months back I tried some experiments with keyword stuffing (well, spamming) for both H1 and H2 tags -- it got me no positive results and no penalties I could see, either. I don't usually try to spam, but the temptation was too hard to resist, once I realized that I could use css to make even an H1 tag into a really small font.
Right now, I'm assuming that any SE which writes H tags into their algo is assuming they are used in a standard way -- such as on the "window blinds" page. So even when I make the font size small, I keep the length of the tag to 12 words max, and then plug major kw's into H1 and H2 tags, and secondary ones into the smaller tags. I've also given up on using H tags in page footers and other obscure spots. It didn't seem to make any difference at all.
Have you ever tried / monitored what effect using only 1 heading over say using 5 has on different engines? Given that if you use 5, you may want to make them natural (ie not repetitious?).
I also have a whole site that is doing pretty well in a competitive field with no H tags at all. So, obviously they are only one tool we can use, not something that is required for good rank. I only started playing with H tags a couple months ago, so the jury is still out for me, I'd say.
To have a good test, I think I'd have to use two identical pages, with only the H tags being different. That would be the only way to isolate the effect, if any, of the tags from other elements.
Given the growing mood at the engines against dupes, I don't think I'm going to try that ... I don't have a page I can afford to sacrifice.
Can you think of another way to test?
For example, instead of this:
< h1 > < center > Webmaster World < /center > < h1 >
could you do this?:
< h1 > < center > < font size=2 > Webmaster World < /font > < /center > < h1 >
(I'm using spaces because I'm not sure if this program will create actual HTML). This is really key in my web page design.
Edited by: Bradley
Like I say - just a theory and I haven't had time to test it recently.
H2{font-size:12pt; color:#112100}
You can define the appearance any way you choose, even with a normal font-weight instead of bold. All that shows on your HTML page is a very straight forward looking set of H tags.
I've been doing this for a while, and it seems to work pretty well. It helps me get some clout for pages that used to be ignored, without blowing the aesthetic the client wants to see.
Just for insurance I place my .css files in a directory that robots.txt disallows, but even before I did that, I never had a spider ask for a css file.
It sort of confirms (I think) what I thought (to be sure - no Irish punn intended).
But what of this putting .css files outside relative to the suggestion Tedster made ( H2{font-size:12pt; color:#112100} )?
CSS can be a complex study, complete with lots of quirky browser support to test through, and so on. The HTML Writer's Guild [hwg.org] has a pretty good FAQ on their site, if you run into coding questions.
Here and there on the web different people have collected lists of bugs and quirky support for various features of CSS, but I haven't found any one authoritative source for bugs and workarounds so far.