Welcome to the Q&A Forum

Hi guys! Thanks so much for the response! I did think it was an algorithm change, however the problem is- overall traffic hasn't dropped, and if anything, like that blue line, has increased and stayed steady. I attach screenshots of search console and mobile traffic. I'm having trouble finding any search algorithm change that would cause a drop in pages receiving traffic but not overall traffic? 6TUJlcR ZO8xe5i

Hello! Thanks a lot for your feedback and clearing this out! It worked well. The robots.txt tester is a good tip! Thanks!

Its a travesty to compare moz with small software like Travis . Moz has Backlink checker with world known metrics PA(Page Authority) ,DA (Domain authority). Web based power which is highly greater than the software one. Rank checker and so many tools that Travis cannot even imagine. Now let's see what traffic Travis can work for one-time payment and few number of tools. Way way cheaper then monthly subscription and not even $ 97 as you said but it's available for around $ 37 on some coupon sites like http://www.discountmycart.com/ so it is quite funny to compare both! travesty

Hmm I don't handle the PPC portion, but that is probably what it is.

Yes those tools are good and the comment about intent is right on the money too. [Apple might have a high search volume, but that's not necessarily as helpful as you might think, if you are a fruit seller.] One thing I like to do when looking at various keywords is to put them in Moz's keyword tool and see which sites are ranking. This tells you something about intent (the sites that rank highest for a search term are likely to be the most closely aligned with user intent) and about competition (the tools shows you some metrics about the ranking domains). It doesn't help to pick a term with a high search volume if the competition is way out of your league.

Hi Jason. While it is a best practice to use the keyword in the URL, I think you did overdo it and are running into something that would be considered keyword cannibalization within your own site. While an old blog post, the core principles remain the same here: https://moz.com/blog/how-to-solve-keyword-cannibalization.  To get around using 'physical therapy' excessively, you can break down the different types of PT into: treatments, exercises, stretches, and so on.  This semantic use of similarly related keywords is a positive signal versus the negative of keyword stuffing.  Cheers!

Happy to help!

Thanks. I'll look into it. We're a journalism operation so we're going to write the stories we need to tell rather than just following the search opportunities. BUT, once we've started to write the stories, I'd just like to provide the writer with information/suggestions that might help their stories do better in search.  And, a lot of what we write about is very, very timely so we'd need to be pulling from almost realtime data.

Hi Patrick, That is super useful and thank you. I have read up as suggested on MOZ and Webmaster tools and you are bang on - absolute over relative urls to be used in canonicals. I will ask our developers to fix. I wonder whether MOZ will ignore this as an error in its crawl diagnostics thereafter? Ben

Ok, well that would not keep me up at night!

...if a serious amount of content on our site was to change, even if the new content was better, more informative and uses better keywords for our site, how dramatically could it affect our rankings? For better or worse... There is no way to tell for sure, but if the proposed content is better, then in theory, you should see positive movements. That said, don't ignore the user experience. You want people to perform an action on your pages, so perform some user testing or heat-mapping (Crazy Egg) to see how people are using your pages. This can often be a huge eye-opener because the worst person to ever try and evaluate how good a page is, is yourself. What you think is good is not always what others think is good. Tracking clicks from the mouse can tell you a lot about where key components of the page should exist. -Andy

Thnak you all for your insights. I've learned a lot!

The problem is that you are implementing a canonical from all those parametrized pages to http://mathematica-mpr.com/news/, and the content in that page -an empty result list- is not the same as in the original link you provide in your post -which show article results. Canonical is used to show a preferred URL for a page, when there are two or more URLs which lead to the same content. After implementing the canonical, you have to make sure you are not linking to the "bad" url or it will never disappear from Google's SERP. As your content is not the same on both pages, and the old URL is still accessible from your site, you are not fullfill the requirements to make those URL disappear. I think that if you don't want those index to appear on the SERPs as depending on the search options it will lead to very common pages, and it is almost impossible to identify clearly a canonical reference for them, the best option is to add to robots.txt a line which blocks those parameters, and you will end with that problems. Disallow: /news/?facet=* This will remove your search results from the index, ending with duplication problems. Please make sure this is what you want, or if you prefer to keep those results although you continue having duplication issues.

I would indeed disavow like you are already doing and keep your lists up to date with the data from Google Analytics, Moz, Majestic SEO and the other tools that are available. They'll keep you up to date on new links coming in, mostly GWT is running behind in doing this.

Thanks to those that posted and emailed! I'll be reaching out in the next week to set up some time to chat.

Hi there I want to say that it may have to do with your canonical tags. For instance... http://www.craftcompany.co.uk/pme-rose-leaf-veined-plunger.html has a canonical tag for http://www.craftcompany.co.uk/pme-rose-leaf-veined-plunger.html But, if you use http://craftcompany.co.uk/pme-rose-leaf-veined-plunger.html has a canonical tag for http://www.craftcompany.co.uk/pme-rose-leaf-veined-plunger.html?SID=1c501bb25ab64ab687f30b714dee9969 Your non www. URLs seem to have a SID parameter attached to their canonical tags. I would check out Google's resource on duplicate content. There are links to help you clean these up. I am not saying that's the reason, but I would definitely clean those URLs up in canonical tags. Hope this helps! Good luck!

Hi there 1. Yes, and go through a backlink audit - what needs to be removed / disavowed? 2. I used this combination of information (as well as Majestic / Ahrefs) for multiple audits / manual action clean ups 3. You can use tools like LinkRisk to assess your backlink profile and score individual links. All in all - the best way to define a bad link is to use your own intuition and common sense: Does this link help my website? Is this link relevant to my website? Would I trust this site (that's linking to me) if I landed on it? Is the website or content in which I am being linked from topically relevant to my website? If you check metrics - does anything about the metrics (domain authority, page authority,Majestic, SEMRush traffic/ranking data, etc) make me feel uneasy? Hope this helps! Good luck!

Thanks Dan!

Hi Christina, We tend to agree with the others - you get what you pay for. Another important consideration is what you are including in your reports and what metrics are important to you. We have always felt that including analytics data into reporting is crucial for setting benchmarks and being able to see the historical data for trends and analysis in one place is also very useful. Many of the free sites can be an ok solution for identifying optimization ideas and opportunities for improvement but lack substance in benchmarking and comparison tracking with M/M or Y/Y true data. We like to use Moz, SEM Rush, and Screaming Frog for some of our day to day work and use a combination of Conductor and Adobe SiteCatalyst (Omniture) for reporting. Hope this helps!

That must be really frustrating for you Mike. I am not surprised you can't access them at the back-end as these are almost always a by-product of the e-commerce system. So many don't care about what extra churn the sites produce. You could always robots out those bad pages? Just something like: Disallow: /store/pc/www.ocelco.com/* This should still readily allow everything else to be indexed after http://www.ocelco.com/store/pc/... Give it a try and then run the site through Screaming Frog and see if these bad pages are then picked up. -Andy