The answer to this questions kind of depends on the technology you are using.

(X)HTML 4 and below will only allow you to use one h1 per page - so it's got to be the most important thing on that page. Think of it like the headline on the front of a newspaper. Your H2 elements are then the sub-heads, h3's sub-subheads and so on; like the newspaper you can have as many of those breaking up the text as you want. This is why you may have heard about people using an h2 element for their main headline, especially on home pages; the most "important" thing there - to the client at least - is the logo, which is therefore often wrapped in an h1.

HTML5, on the other hand, lets you have multiple h1 elements (one per section). So you can wrap your logo - up in the header - in one, you can have one for your page's main headline - it's in a

or-  and your sidebar can have one - it's in its own <sidebar>after all. You can even have one in your footer if the whim takes you. Search engines establish which h1 is more important through a hierarchy (article is more important than sidebar is more important than header or footer).

Hope that helps!</sidebar>

This was a question that was recently raised by somebody else here on the Q&A and you can see the full discussion here. Although the question is slightly different, the answer is the same: it depends which HTML version you are using. If it's HTML4 or XHTML, only use one h1 per page, but if you are using HTML5 you can have one per section. So you could have one in your

<nav>, on in your, one in each of the

<aside>s on your page and one in your

<footer>.

The reason for this is that for HTML4 search engines look at headings to give the page hierachy - remember they only had

s to separate content areas - but HTML5 uses the new semantic elements like

<header>and <fotter>to work out hierarchy, with headings only affecting hierachy within one of those tags.</fotter></header>

</footer>

</aside>

</nav>

I guess this is less of a question about SEO for HuffPost and others than about reader engagement - I'd be fascinated to see whether anybody who runs a news site or blog with a large number of page views (cough Moz cough) has data on this.

Going with my gut, I would say that longer headlines give the reader more of an idea about exactly what they're going to read, so they are probably more likely to engage with the post than if you used a shorter of headline. Of course, with bounces and other engagement metrics potentially being part of the search engine algorithms that might give you a bit of an SEO boost.

i wouldn't think it would be an SEO problem to have a different header on your subpages to your homepage. There are some companies who have done this to good effect. Dropbox is a good example; they have no navigation on their homepage but do on other pages. There are also plenty of people who change their headers on landing pages and during the checkout, to remove the distraction of a menu.

However there may well be development problems, depending on what platform you're using. And from the users point of view, what is the benefit of having different navigation in different places on the site? Or are they likely to get confused?

Search engines look at the content on all of your page, so comments will count as well. Get lots of engagement on a short article and you've got as many words on the page as if you'd written a short article and got no comments.

I would tend to agree with Chris though. Thinking about quality, engagement, and relevance will get you much more in the long term than just writing long articles that don't engage people.

I'd agree with all of the above and say that directories are essentially useless for improving rankings, except when they count towards local citations.

Having said that, I would suggest getting a listing in any directories that rank for search queries you want to appear for. That way even if you don't get clicks directly through from your website being in the SERP you might catch a few customers coming through the directory that you otherwise wouldn't have.