Rankings Continue To Drop

[charlesfitz]

Hi there

I'm at wits end trying to stop the slow bleed in our rankings to our store URL's that started mid March 2017 and continues through to today. I'd appreciate some pointers and hope this will throw up a challenge to someone out there. Here is the background:

1. We run an e-commerce store on Shopify with a blog. The recent ranking decline has been almost entirely on the store URL's (catalogue and product pages) while at the same time we have seen steady growth in search volumes in the blog - this makes me think we are seeing a Penguin4 penalty of some type, because the impact is confined to the store URLs.

2. We received a linked based manual penalty back in 2014 and this was successfully removed within 3 months. We have quite a large disavow file as a result.

3. Shortly after launch of Penguin 4.0 in Sept/Oct 2016 we saw a really nice boost in traffic and ascribed this to being under a previous Penguin algo penalty, now removed.

4. Come March 2017 we see a small but steady weekly drop in rankings for our store URL's only, this steady drop continues through to today and over time has become significant. Approximately a 50% decline in visitor numbers to store URL's only as of today, since March.

All of this despite:

a. Initially I thought this was a Panda issue (because it seemed to coincide with Panda releases in March and May) so the entire website has been rewritten (during June and July) with thin content removed across the store and the blog. Remaining content has been given a serious content boost, being very careful to watch for over-optimisation, and for keyword cannibalisation. I think I've got this right. There are also no crawl issues being highlighted by Moz Pro or SEMRush site audits.

b. Recently discovered, only last week in fact. A very low domain website, trust score (0 and 0) had been copying our blog articles steadily on a weekly basis, starting Oct 2016 (yes same time as Penguin4) and only caught last week (my fault for missing this). These articles were copied verbatim with all links and so generated nearly 400 spammy backlinks to our store URLs (about 30% of all the links we have). I've had all these articles removed from the spammy site via DMCA so none of those links exist anymore (as of 8/14/17). I've also disavowed this domain with Google. Could these spam links be the issue, and Google is still needing to crawl this site to see the links are no longer there? I'm not sure because my understanding is that Penguin4 would have devalued these links to start?

c. A general review of links and anchor text. I've used Moz Pro and SEMRush backlink audit (linked to Google Search Console) and have removed all toxic links by contacting web masters and using Google disavow. This included removing any links that I think are causing over optimised anchor text. After disavow, according to SEMRush, we have no toxic backlinks left and only 50 out of 1200 links with "Money" anchor text. This exercise was completed two days ago when the last disavow file was uploaded. However I don't believe there was an issue here before as toxic links were < 1% of all links and exact match "money" anchor text in the region of 5%.

d. One potential problem with our backlinks is that we have quite a few high domain/high trust links to our scholarship page with anchor text "official website". The net result is that our "Other" anchor text category is just over 50% of total links - these are mainly educational institutions with .edu domains.

e. A review of internal linking. We had some what I would refer to as SEO links, linking all product and collection pages across the store, through a tagging type system. This was removed two days ago as it was probably unnecessary for user experience. Other than this I have two concerns remaining with our internal linking structure. The first is that we have quite a big static navigation on the left margin of our store collection pages. This is not faceted navigation, but static. The second is that we've internally linked from almost every blog to our "key" money page in the store, however with varied and non-money anchor text.

f. There is nothing in Google Search Console indicating a problem, no manual actions, no significant HTML improvements, and Google has indexed over 90% of URL's compared to the sitemap.  All broken links have been fixed - there were a lot before but all fixed as of three weeks ago.

g. Checking site speed in GA. Speed has remained constant over the period and we have put in some fixes to improve it. Site speed has not got worse and scores average in Googles speed checker.

That's about it. It's possible that with the recent changes made with respect to b, c, e and f above I just need to wait a couple more weeks for Google to catch up, and would appreciate thoughts on this.

However I'd also like some thoughts on the static navigation on our collection pages, plus importantly on linking from blog articles to mostly a single money page in the store - of all that remains I think this is potentially a problem.

Our website is located at www.thekewlshop.com

Many thanks for your help.

Charles

[DonnaDuncan]

Charlesfitz,

You've done an impressive load of research. Here are a few other items to consider:

1. 464 pages had a connection refused (0) status code. Here are some examples:

   /collections/nude-mesh-dress/products/so-sexy-see-through-bandage-dress 0 Connection Refused
   /collections/bandage-dresses/products/navy-blue-cap-sleeve-bandage-dress 0 Connection Refused
   /collections/mermaid-midi-dress/products/mermaid-bodycon-bandage-strapless-sheath-dress 0 Connection Refused
   /collections/sexy-birthday-dress/products/celebrity-bandage-pink-v-neck-dress 0 Connection Refused
   /collections/whats-new/products/rio-rose-gold-two-piece-bandage-dress 0 Connection Refused
   /collections/whats-new/products/bandage-black-strapless-pencil-dress 0 Connection Refused
   /products/spaghetti-strap-lace-dress 0 Connection Refused
   /products/hot-celebrity-black-and-grey-bandage-dress 0 Connection Refused
   /collections/whats-new/products/two-piece-lace-bandage-dress 0 Connection Refused
   /sexy-black-dresses/products/black-bandage-spaghetti-strap-midi-dress 0 Connection Refused
   /collections/one-shoulder-dress/products/one-shoulder-gold-bandage-dress 0 Connection Refused
   /collections/white-bandage-dresses/products/white-elegant-strap-bandage-dress 0 Connection Refused

2. You have roughly 45% of your HTML pages returning a 430 status code. I have no idea what that is but it might be impeding your indexing.

3. You have 80 temporary redirects in place for HTTP to HTTPS files. That's not likely the explanation, but should be fixed.

4. You have a lot of duplicate title and H1 tags. That might result in the wrong pages getting ranked, and by wrong I mean the pages that aren't ideally optimized for your keyword phrase or targeted for SEO equity accumulation.

5. There are 13 404 (page not found) status codes.

6. You also have some perhaps unintentionally noindexed pages, for example, /blogs/news/top-10-new-york-fashion-bloggers, /collections/gold-bandage-dresses and /collections/white-swimwear. (It might also be intentional as I notice you can't manually navigate to white swimwear on the site.)

I found these by scanning the site using Screamingfrog.