Topics such as the value or importance of h1 tags set apart capable SEO minds from the hordes of people who insist on the value and benefit of headings and other text styling aspects of documents crawled by search engines. Take a moment to think like a search quality engineer. What ranking factors receive the highest priority? The answer is quite simple: those factors that are not easily manipulated by people who create documents primarily in an attempt to gain position in search results via easily established means, especially those means that might relate directly to the markup or code of a document. Secondarily, one might consider that to incorporate such a factor in the measurement of a document’s value and rank, several things must happen: collection, storage and measurement of the data being the primary activities around these markup and text elements – all of which contribute to increased machine resources. Every factor taken into account requires a reference and every reference requires storage. Every calculation requires processing power, and many factors would require that particular strings be parsed out of the code and incorporated into the established calculation or calculations for rank. So, with even a small consideration factored into the mix such as an h1 tag surrounding some string, the increase in resources would be significant when placed on a scale such as Google’s. That said, it should be quite safe to assume that any factor incorporated into ranking algorithms must be a signal capable of pointing to a document’s relevance or quality. And, such a signal must be efficacious, measurable, and probabilistic.
Where then does an h1 tag fit in all of this? It doesn’t.
CSS and Document Styling
What CSS did was allow the h1 tag to be stylized in a manner that was not the grossly overwrought style that was native to the pre-CSS h1 tag. The default h1 displayed in an obnoxiously large font size (write some style-less HTML to see for yourself), which few if any commercial sites would dare put before an audience – a style used almost exclusively in situations that were academic in nature (whitepapers and the like). Those that would use the h1 tag for the sake of optimization were (without many extra steps) bound to the default style in the early days of HTML, and being that there were a dozen other, easily employed methods of optimization that were available and effective, only a hungry SEO would put an h1 tag to work. This is exactly the reason the h1 tag played a role in rank: it wasn’t easily manipulated (pre-CSS) and would likely constitute a strong signal for serious content.
Remember to continue thinking like a search quality engineer, and imagine the implications of CSS for any and all page styling that might have been given weight as a content quality signal. Every bit of that is completely washed by the capabilities brought to the game by CSS. So what do you do as a search quality engineer? You realize that few signals are left that could come from within a document and be deserving of variable status in calculating rank, because those signals can (with CSS) easily be tweaked to throw your algorithm a positive, but invalid signal. This includes all other heading tags, bold, italics, underlining, emphasis, and the whole mix of what may have once been a positive quality signal (if any signal at all).
What the Experts Say
When I see h1 tags given a score of moderate importance in the SEOMoz search factors vote – a culmination of opinions from 72 “SEO experts,” it’s clear that there are plenty of “professionals” who practice SEO by way of reading and accepting non scientific pronouncements as fact and never stopping to question or even think that the very idea is by nature incredibly unlikely. I wonder how many of these pros can write HTML or CSS without heavy dependence on a reference? No one should be allowed to vote in this event without proof that they are able to show a set of core competencies, such as a comprehensive HTML aptitude. I would even go so far as to say a truly qualified SEO should have a measurable penchant for text processing, computational linguistics, statistical analysis or at least some ability to act as scientist or in a strong technical capacity, despite the limitations at hand. Being an English major is not enough. Reading blogs is clearly not enough. And, if you’re going to be stumbling in the dark as most SEOs are, you should at least have the wherewithal to draw conclusions based on logic. You should know that at the core, Google’s algorithms are rooted in mathematics and probabilities derived from the frequency at which words/strings appear inside and outside of documents. And those outside of documents determine the probability of authority and quality.
SEOMoz further expands the sad truth with these details:
The primary heading tag (the h1 tag) ranks 4th among all surveyed, on-page keyword factors. According to the experts, it ranks above all of the following factors:
Keyword Use in the Subdomain Name (e.g. keyword.seomoz.org)
Keyword Use in the Page Name URL (e.g. seomoz.org/folder/keyword.html)
Keyword Use in the Page Folder URL (e.g. seomoz.org/keyword/page.html)
Consider this question: Do you ever see the contents of h1 tags displayed overtly in search results?
You do see, however, the keywords in a search query that match any of the (mistakenly) “lesser” URL-based factors highlighted in search results – as you find them in bold. Would this then perhaps indicate that there is some bit of weight given to these? If not a factor in rank, these portions of a document, are front and center to indicate any keyword matches to a query. So, if nothing else these highlighted strings have the potential to improve click appeal. Meanwhile the h1 tag inhabits the cruft bucket, it’s highest honor is simply that of being cached somewhere. Hooray for h1!
How can someone claim to be an expert if they never stop to think about SEO like a scientist? The problem with SEO knowledge is that it’s primarily anecdotal – passed along indefinitely until someone approaches it empirically or there is a direct counter position from an authority. When you have sites such as WooRank and Website Grader perpetuating the h1 mythology, it’s certainly not likely to go away soon, even it someone disproves the value of h1 tags in SEO.
These sites aren’t there for the sake of science. Such sites came to be based on the value of linkbait. The problem is that these “analytical” sites are so far from the heart of what SEO is really about, that they might as well be dog shows. This is good for search results quality, however – as the masses do put faith in their “authority” which is nothing more than the perceived need to create a longer list of considerations so as to appear to be the most comprehensive “tool” among tools. The real rabbit and the real hat are elsewhere in a place where content is created for the sake of sharing knowledge – real knowledge – and not some stylized, value structured concoction more concerned with commission or some rehashed marketing conquest.
Now – does this mean you should not use h1 tags? By no means am I saying that you don’t use any of the style elements that I call out here. They all have a place in markup and many are essential to typography (an afterthought for most, unfortunately). It’s just that they don’t have a place in SEO. They haven’t for a long time. Headings are simply tossed.
And, one would only hope. How abysmal would search engine results quality be otherwise?
But, please do use heading markup for your visually impaired site users. They depend on your thoughtfulness for such details. Surely you care more about them than you do manipulating search engines, right? Put yourself in the position of a visually impaired site visitor and then ask yourself if your headings markup is in good order and able to help you make the most of a page you can’t process sufficiently by sight. And, if you’ve been using h1 tags to tease the spiders, maybe you are impaired in your own way.
Follow up: I found this on SEOMoz, shortly after writing this article:
The Usefulness of H1 Tags
H1s are important for website visitors but not necessarily for search engines.
Our correlation data shows that H1 tags do not carry the same ranking weight that we had originally presumed. We think they are very important for establishing information hierarchy and helping with algorithmically determined semantics, but they seem to be less important for search engine optimization. We recommend them on all pages as an aid for users but don’t stress the importance when other opportunities for SEO improvement are available.
Here’s another great resource on this topic: