SEO+ by embedding H1 inline?
April 21, 2010 10:13 AM   Subscribe

I've got an SEO consultant telling me to inject H1-tagged phrases into copy on a website. Is there any evidence that this either works or hurts?

This thread describes the same technique that I'm being advised to deploy. The idea is that you leverage the extra weight of the H1 tag to emphasize keywords for the benefit of search spiders. For example, SEO might be emphasized by wrapping it in an H1 tagset (as "<h1>SEO</h1>"), then re-setting H1 so it renders inline instead of block (h1{size: 1em; weight: normal; display: inline;} or something like it).

To be clear, this is NOT about creating semantically structured web pages. I'm aware that the resulting pages will be semantically broken, in fact, since they will be littered with top-level headings that break up the natural outline of the page. Also aware that this poses some significant technical issues for CMS-driven sites. (Basically, means you can't filter for broken HTML anymore.)

My question isn't about whether this is a good or bad technique semantically or in technical regards apart from SEO. I know it's bad in those regards. This is about whether anyone knows of evidence that it will help, or that it will hurt. What I've read so far is all opinion, no metrics, and no one even claiming actual metrics -- everyone's all about subjective evaluations, as far as I can see. (Personally I hate this idea. But I want to know what the current scoop is on this. Especially whether this tells us anything we ought to know about this particular consultant.)
posted by lodurr to Computers & Internet (16 answers total) 19 users marked this as a favorite
 
Hint: Just about anything a SEO consultant tells you is going to either be ineffective or will hurt you.

SEO tips that work:
  • Have interesting content that is updated frequently and that your users will read, use, and link to.
  • Structure your content semantically so that it's easy for machines, humans, and humans using machines to read.
Fin.

Anything else is tomfoolery, and search engines downrank sites who participate in tomfoolery as soon as they catch on to the tomfoolery.
posted by SpecialK at 10:17 AM on April 21, 2010 [12 favorites]


SEO is snake oil. By the time the search engines downgrade you for doing just these things, the SEO guy will be long gone with your money in-hand.
posted by rhizome at 10:22 AM on April 21, 2010 [1 favorite]


What we mostly go to SEO people for is keywords. We have one guy we've used a lot over the years who's very good, no "tomfoolery" -- he's all about clear, keyword-rich content and microcontent (especially human-readable URLs). But he's so busy we have to look elsewhere. So this is the first comprehensive recommendation we've gotten from this company, which is a more traditional media-buying outfit that has an SEO practice. They've been in the SEO business for about 5 years.

I'm finding references to this technique going back to about 2005, FWIW, and all the evidence either way is based on gut feelings. No one's citing an A/B test on it.

In lieue of ##s, I would be happy to hear whether the technique is penalized, either explicitly or implicitly. E.g., I hear rumors that there are penalties for multiple H1s on a page, but again, nobody can cite chapter and verse.
posted by lodurr at 10:25 AM on April 21, 2010


There's white hat SEO and black hat SEO. I do strictly white hat stuff and have managed to obtain the #2 position for a certain keyword in just a couple of weeks after entry.

So how do you tell the difference? My heuristic goes like this: if I were a programmer at Google trying to get rid of linkfarms, spamkings, pornlures, and other scumbags, and I saw this page, then looked at the source ... would I reward it or would I punish it?

Google always catches on, eventually. If they catch on and you have been naughty, either the method you have been using goes from a positive to a negative and your site takes a hit, or your site gets on some myserious blacklist.

Google has already caught on to various tricks involving CSS. I can't imagine that this one has slipped through their nets. Nobody other than a handful of people at Google could let you know, and I doubt they would. You can only operate by inference here, I am afraid.
posted by adipocere at 10:29 AM on April 21, 2010 [2 favorites]


So how do you tell the difference? My heuristic goes like this: if I were a programmer at Google trying to get rid of linkfarms, spamkings, pornlures, and other scumbags, and I saw this page, then looked at the source ... would I reward it or would I punish it?


Yay, I knew I couldn't be the only search marketer on Mefi.

Our golden rule is pretty simple:

Google isn't stupid, so don't treat it like it is.

If you're looking for hard evidence about this, I suspect it will be hard to find. It's the kind of tactic that fair, open search markers, (the kind who actually collect evidence and have a methodology,) stay away from.
posted by generichuman at 10:34 AM on April 21, 2010 [1 favorite]


This idea sounds ridiculous. We're currently working on a few SEO optimization projects where I work right now and while having h1 tags on pages is important, good and relevant content is way more important. We have been given advice to rework our existing h1 tags but when we try and get an anlylsis of how it will affect our ranking (many of our pages with "sub-optimal" h1 text are in the first five results on Google) we're just told that it's a matter of best practice.

In my experience good SEO is a combination of things that make the page useful to others, and secretly embedding h1 tags definitely does not fall into that category. Since the practice brings with it a score of other issues putting in the effort to implement something that might make a slight difference not really worth it. Your energy is probably more useful focused on things like keywords and linking strategies.
posted by Kimberly at 10:41 AM on April 21, 2010


Answer to Your Question
It sounds like you're talking about using multiple h1 tags, which is a bad idea. Especially for wrapping them around paragraphs. Here's Google's Matt Cutts affirming this: http://www.youtube.com/watch?v=GIn5qJKU8VM

Generally, using keywords within HTML markup that's meant to emphasize the important text on your page is a good idea, so long as you're not abusing it. Most SEO practitioners agree that the <title> tag is one of the most important places you can put a keyword: http://www.seomoz.org/article/search-ranking-factors

A quick comment on SEO
If you think of search engines' jobs as helping people find the best content based on the few words they enter, you might think of SEO practitioners' jobs as figuring out what factors search engines use to determine how good content is. That's not snake oil; that's useful information and common business sense.

If you use that knowledge to try and wrap crap in a candy bar wrapper, you're still just selling crap, and you won't be able to for long. If you use that knowledge to create a better candy bar and package it better, you'll sell a lot more candy bars.
posted by ElfWord at 10:45 AM on April 21, 2010 [10 favorites]


As I think this through, and in reading generichuman's & adipocere's responses, the hardest evidence I find may be in the breach: the technique is only described in a few places, and when people move to debunk it they're often misunderstood as advocating against using H* tags at all -- i.e., it's sufficiently fringe-y that when people have it described to them, it doesn't register and they think the person must really be describing something else.

The other recommendations were pretty non-controversial (set up VERIFY tags for major spiders, changes to the copyright tag, etc.); this one kind of threw me for a loop, as you can see.
posted by lodurr at 10:45 AM on April 21, 2010


elfword, thanks, that's as close to a smoking gun as you get ;-).

The "crap in a candy bar wrapper" analogy would go over well, here. (Which is one of the reasons I'm still working for a marketing firm: this particular one is a little less deluded and a lot more ethical than most.)
posted by lodurr at 10:50 AM on April 21, 2010


I created a site for a client's small business (niche, but there are plenty of competitors)-- this site comes in at #1 more or less globally (i.e., google.com/.ca/.nl) for around 15 top keywords, and is in the top 3 for about 20 more. Site traffic increased threefold when I took over, and all I did was the following (and I can't attest to the true value of each of these, in terms of SEO-- but here's what worked for me):

1. Made sure to have relevant, accurate title tags (no tricks, just titles)
2. Made sure to have relevant, useful content in the main site.
3. Created complementary blog as a repository for useful content that would otherwise clutter the main site. Linked back to main site for those who needed additional help. This has captured a ton of additional search terms and traffic-- instant results.
4. Marketed for free link placement on a select handful of relevant sites; purchased advertising links (cost was about $200 on a relevant site) that serve as backlinks to the client's site-- and if backlinks are no longer relevant or were BS in the first place, no harm done, as actual traffic comes from those links).
5. Created blurbs for Google to use- the intent is to actually capture the traffic by telling them more about the company, rather than just the snippet that Google provides. Have only done this for a couple of the main pages. No idea if it works-- just seems logical to me based on my own searching behavior.
posted by mireille at 10:51 AM on April 21, 2010 [1 favorite]


Seconding ElfWord's comment. I work with two different SEO people through two different clients, and both of them are good SEO consultants who've achieved noticeable results through white-hat methods--structuring content well, tagging it as Google et al. want it to be tagged (e.g., a single H1 near the top that actually identifies the main header in the page), etc. They bring knowledge about how Google et al. handle things and how to measure what the search engines are looking for.

There are lots of black-hat SEO guys out there, and it's tough for a client to tell the difference, though. The rule of thumb for a reasonably knowledgeable client is that, if it sounds like common sense, it's probably white-hat; if it sounds like a trick (as the OP's technique obviously is), it's black-hat, and will bite you in the ass when you're caught using it.

Generally, if an SEO guy is telling you to just make your site good and trust Google, he's probably giving you good advice that actually works. If he sounds like he's trying to get you to game the search engines, find a new SEO consultant.
posted by fatbird at 10:56 AM on April 21, 2010


Wow, this far and no one's mentioned website accessibility? (I watched the video elfword links to and though he does mention the general rule, he doesn't bring up how far accessibility guidelines go concerning it.)

h1 tags are header tags; headers should only be used for headers. Why? Humans who have functioning vision are not the only ones who "read" the web. WCAG2.0, guideline 2.4.6: Headings and labels describe topic or purpose. Browsers for the blind or vision-impaired will read h1 tags as being headers, no matter what visual style you give it to fool sighted readers. Blind and deaf users may well have a Braille keyboard, which will "translate" h1 tags the same way.

WCAG are the Web Content Accessibility Guidelines, a W3C recommendation.
posted by fraula at 11:06 AM on April 21, 2010 [1 favorite]


Wow, this far and no one's mentioned website accessibility?

Because the rest of us read the question:
My question isn't about whether this is a good or bad technique semantically or in technical regards apart from SEO.
posted by juv3nal at 12:36 PM on April 21, 2010


Followup: We talked with the consultant today. She'd outsourced it to someone else in her office, and promised to forward his rationale. When she did, what we got was just a partial explanation of how the stylesheet would work, along with an acknowledgement that it would be of marginal benefit in any case ('don't bother doing it if it's difficult to do'). So they just assumed it was a good thing to do -- they don't really have any supporting evidence.

So, this just doesn't happen on our sites. What we may do is use a STRONG tag with a class that sets the font-weight to a lower value (say, 500 or 600), so people don't get visually hammered by all the emphases.
posted by lodurr at 1:52 PM on April 26, 2010


Final followup: We gave feedback to the vendor, and heard back from the guy who'd done the actual report. He totally ate crow, said this was a trick he had "in his bag" going back a number of years, and that after due diligence he admits it's not valid and won't be using it anymore.

This was good for us to hear because everything else they gave us was common-sensical. We don't have the bandwidth to insource it, so we need to have people we can rely on -- we've used them to outsource adwords and print/broadcast placement in the past so we have an existing relationship.

One lesson I take from this: If you have concerns with how a vendor is doing what they do, make sure they know your concerns and trust your own judgement about whether the methods are valid. They may not really examine their methods unless people make known that they don't like them.
posted by lodurr at 7:05 AM on April 27, 2010


[final-final/aside: I have reason to believe that adding VERIFY meta-tags had a dramatic impact on uptake speed with Google and Bing. YMMV.]
posted by lodurr at 7:11 AM on April 27, 2010


« Older Make an online list of books I've read?   |   Clicky McClickerson Newer »
This thread is closed to new comments.