Identifying a Negative SEO Campaign

[irvingw]

main keyword anchor text is sometimes not exploited with a negative SEO attack, it's just a massive amount of links from bad sites which harm the site in general. this can easily be detected with link detox for example or you can do a link:(space)www.yourdomain.com search in Google.

[Dan-Lawrence]

Thats a great answer Robert, i really appreciate you taking the time to comment so helpfully

I should have added that there was a big rise in backlinks beginning of may, that peaked and levelled throughout June to then drop from beginning of July to date (according to ahrefs data). So in an otherwise nice natural looking link growth rate from nov last year to date there is a huge hump or wave in the graph as links rise in may but then drop over july.

So if i was looking into this in June it would, initially at least, look like it could well be a neg campaign, but the ranking drop has only occurred recently, correlating with the drop, not the rise, in links.  If a neg campaign i would have thought the rank drop occur soon after the spike in link growth, not after a drop in links.  Also the link growth period is spread over a month (as is the period of the link drop too), not a few days as article suggest one should look out for in a neg campaign, hence i'm pretty confident that its not (which is why i didnt mention it originally but thought best to now just in case).

When you say look at CTR do you mean purely in regard to traffic from the effected kw in the run up to the rank drop ? What kind of time period do you recommend, a week or more ?

Cheers

Dan

[Dan-Lawrence]

Thanks Irving

do you know if any free versions of linkdetox ?

how will doing a link:(space)www.yourdomain.com search help since results wont highlight site quality will they ?

All Best

Dan

[Dan-Lawrence]

Also does anyone know which is the more accurate data source MajSEO or Ahrefs since im getting wildly conflicting data from both.  MajSEO now showing 845 links added for early July (which would indicate a neg link campaign) but hrefs shows 345 links lost over same time period ! ?

[nsauser]

Could it be something else, like a Panda update?

I agree, a typical negative SEO campaign in my mind is a ton of easy to acquire links. I doubt anyone is going to take the time to email webmasters and have links pulled.

I would look at your content stats in GA for YTD and see if you can see any trends for the pages that lost rank (or was it the homepage?).

Unless the negative campaign is targeting individual pages then I would assume the whole site would be affected.

[Carson-Ward]

Most claims of negative link efforts are either not link-based penalties, or something self-inflicted. I'd be fascinated to see one, but it doesn't sound like that's what's happening. This is the process I follow in diagnosing penalties:

The very first step is always to check webmaster tools for messages - or now you can check for manual penalties. Second, look at the date of drop and compare it with http://moz.com/google-algorithm-change. Try to figure out when the drop was and whether Google was making any updates at the time.

Link profile: look first for overly-targeted unbranded anchor text first, as you did. Don't forget to pull the link info from Webmaster Tools and check for newer links from Moz newly discovered or ahrefs.

Panda or other site/page quality updates: If you couldn't tie it to an announced Panda release, we just have to guess. Is there a heavy template? Are there a lot of pages targeting very similar terms? Is there any "form-letter-like content"? Is the organic bounce rate/time on site very bad?

Link profile part 2: Look through the linking sites. Check for  looking for links that are clearly ads, but lack the nofollow attribute. Start with sites that have been knocked down to PR 0 despite having plenty of links, and look for paid links, especially of the site-wide or overly-targeted variety.

Finally, remember that rankings can fluctuate without it being a penalty.

Google might suddenly "realize" that "Cavendish" is a biker in addition to a type of banana, and might also refer to a philosopher. They'll then push more diversity into the SERP for disambiguation, which will cause rankings to fluctuate wildly. (QDD). Sometimes Google just devalues links that were helping you to rank - it's not a penalty, but it has the same effect. Sometimes we just don't know; the rankings might pop back, and they might not.

If you come up empty after all that diagnosis, you have only one choice: carry on building great content, optimize the design and structure for users, and work on building awareness and authority throughout the industry. It will pass eventually, and you'll come back stronger for having built value and done real marketing.

[Dan-Lawrence]

Panda would be site wide not kw specific !

It was the HP

there is link growth of aprox 1200 backlinks (but only an increase of 4 domains) over 3 days according to MajesticSEO (although hrefs reports a link & ref domain drop over same period)

[Dan-Lawrence]

Thanks for taking the time to comment Carson

One of the reasons friend/client has hired me to look into this is because he cant retrieve any link data from GWT. Have asked him to check messages and new manual notifications too.

If Panda surely probs would be site wide not kw/page specific ?  Its the HP thats dropped for a primary target kw when all other similar kw rankings for hp are still page 1

Also interestingly there is link growth spike of aprox 1200 backlinks (but only an increase of 4 domains) over 3 days according to MajesticSEO although hrefs reports a link & ref domain drop over same period, any idea which is the most accurate/trustworthy data since directly contradictory date, one of those data sources must be WAY off the mark !?

cheers

dan

[Carson-Ward]

Is it possible that there are so few links that losing a few cause a big change in ranking values? Was the keyword in question an exact or partial match with the domain name?

As far as accuracy, it can go either way. Majestic often reports links that existed but are now dead, but it can sometimes be more thorough. The only way to know for sure is to check the links and see if they're still live.

Panda is often site-wide, but it can apply to one or several page/kw combination, or even to all the pages with the same layout in a section.

[Dan-Lawrence]

I would have thought if that were the case would effect the other partial match variation rankings too, which it hasnt since all others are still on page 1 !

The kw in question (along with the other kw still ranking on page 1) are partially matched with the domain