| 9:04 am on May 15, 2014 (gmt 0)|
There are 3 problems:
1. Q&A sites and forums are targets for Panda.
2. Remove duplicate generated pages (delete or use canonical tag).
3. H1 abuse si "classic", remove it IF is not the title. (ex is the content of the questions.
| 9:15 am on May 15, 2014 (gmt 0)|
That's what I thought,
Will removing the H1s again restore traffic or make it worse? (unfair question I know)
Thanks for your insights.
| 10:40 am on May 15, 2014 (gmt 0)|
Is there one H1 headline per page? If so, what problem would Google have? That's semantically correct. If you have more than one H1 per page, then that makes no semantic sense at all, and you're confusing Google.
| 11:05 am on May 15, 2014 (gmt 0)|
Yes only one. And semantically that does make sense but GG doesn't seem to like it. Their algo reacted negatively to it.
| 1:25 pm on May 15, 2014 (gmt 0)|
@ColourOfSpring, my understanding is 1 h1 tag per page but several h2 tags is okay, is this also your view m8?
| 1:32 pm on May 15, 2014 (gmt 0)|
@CaptainSalad2 yes that is correct.
| 1:38 pm on May 15, 2014 (gmt 0)|
CS2, yes - absolutely. H1 is the most important headline (top of the pyramid) - then you can have more than one H2 (below the top of the pyramid), more H3s and so on (on the same page). In reality, most pages might have just one H1 and nothing else, or one H1 and two or three H2s....(typically).
@petra, it makes no sense for Google to punish your site for making each page that bit more easier to understand (from spider perspective). I can only think that perhaps your CSS is making HUGE text with H1s and Google are seeing the text is physically too big (I'm sure they can measure this)? << that's just my outside guess, probably not the case with you.
| 1:52 pm on May 15, 2014 (gmt 0)|
@ColourOfSpring I agree, not sure it's the CSS, as it's not styled to be that big. The issue I believe is that many of the questions are unanswered/or answered with one response that coupled with the duplicate page issue has red flagged it as a possible over-optimization (h1 is similar to title tag). We're going ahead and removing the H1s for now and hoping it will recover the lost traffic. Will report back here with results in the meantime, any further insights very much welcomed...
| 2:52 pm on May 15, 2014 (gmt 0)|
Did it really happen overnight after the H1 tags were added, or was that figurative?
Cause Google doesn't usually act quite that quickly.
| 3:44 pm on May 15, 2014 (gmt 0)|
Has the rest of the code on the changed pages been checked very carefully to be sure a different error didn't accidentally get introduced at the same time (noindex etc...) as the template changes.
I find that changing h1 has much less impact on results than it did a few years ago, and would be very surprised if that was the cause, especially (as netmeg says) if it happpened very soon after the change.
Personally I wouldn't bother removing them again to try and please google - or maybe just remove them on 20% just to be sure - since it is good practice to have a h1 on a page. But I would certainly give priority to sorting out the other issues you mention re duplicate pages etc.
| 3:57 pm on May 15, 2014 (gmt 0)|
My view is that one h1 tag per page is not going to penalize a site, it is normal to have one. Thousands of duplicate pages are something to focus on and fix as soon as you can. What are you seeing in GWT?
| 5:50 pm on May 15, 2014 (gmt 0)|
@netmeg gradual decline over the past 10 days but very clear that it stared happening after the change
| 5:53 pm on May 15, 2014 (gmt 0)|
@Rasputin checked everything else, this is the major change, but will check again.
@not2easy only signal in WMT is the dupe pages, but these have always been there.
| 6:06 pm on May 15, 2014 (gmt 0)|
It is possible that the h1's triggered a review of the site content, which was found lacking and in part, duplicative. If that is the case it is unlikely any recovery will occur. Sometimes fixing something that isn't broken to chase more SEO and ranking just isn't worth the effort.
| 6:09 pm on May 15, 2014 (gmt 0)|
@tangor I agree. Most likely scenario. Do you think it was a manual review?
| 7:21 am on May 16, 2014 (gmt 0)|
This Google patent (against spammers) may shed light on the issue...
Bill Slawski explains...
|For example, the initial response to the spammerís changes may cause the documentís rank to be negatively influenced rather than positively influenced. Unexpected results are bound to elicit a response from a spammer, particularly if their client is upset with the results. In response to negative results, the spammer may remove the changes and, thereby render the long-term impact on the documentís rank zero. |
If this is the case here, then as mentioned previously in this thread, best to ride it out, especially as the change is not an attempt to spam but to add logical semantic structure to the website's "documents".
| 12:43 pm on May 18, 2014 (gmt 0)|
The 'deter tweaking' patent could be responsible; if it is then removing the <h1>s will do no good - at least not in the short term.
I also agree with Tangor's suggestion (that changes to the page triggered a review of the page) as a likely explanation for what's happened.
When Google move the goalposts they do sometimes seem to grant amnesty to sites that are now 'over the line' - so long as nothing changes. Changes mean re-evaluation under the new rules. I have seen this first hand with links. It seems very reasonable to assume that they would apply this approach to content too.
When things are put right, recovery follows (at least that's what I've seen). I think the process either way is not a manual one. Manual doesn't scale.
| 10:00 am on Jun 3, 2014 (gmt 0)|
Hi all, many thanks for your thoughts and input regarding this issue.
The issue was not with the heading tags after all, it was with the removal of the related questions, so overnight, the questions pages lost links from other question pages and therefore became like deserted islands in GG's eyes, therefore dropping out of the SERPs.
The Way Back machine (thanks Alexa!) helped us see what the old pages looked like.
A day after re-implementation of related links, traffic jumped by 10% and another 10% the day after. Now it's back on the increase and heading to previous levels (if not better)
Thanks again all and sorry of the red hearing and many thanks for all your insights and support.