SEO News

Modi’s Moz article is for an SEO & dev lead to implement together, but this guide is more designed for client services to help more concisely communicate the most prescient issues and explain the value in each (and relatively costly) step.

It assumes that you have done all the project planning, timing, and already laid out the processes.

Benchmarking

The key steps of benchmarking are often forgotten in this process and are a particularly key consideration for client services.

There’s no doubt a lot of pressure and high expectations for the new website to flat out beat the old one, as well as the fact that any attention going into the redesign and migration process shines a light on all this.

However, with the best will in the world and all tick boxes checked there will inevitably be a dip in visibility, visits, and the value they generate, so be prepared to have some answers when the time comes.

Benchmarking allows us to show up any legacy issues the previous incarnation of the site may have had, issues that may potential hamper the new site.

We’re talking about factors you can’t measure at a later stage here.

Although the plan will be to iron out any of these issues and to embed SEO throughout the new build it may well be that more focus/time/cost could have been spent on SEO previously, so bear this in mind.

An example might be the use of a clever method of serving content such as frames or dynamic divs in a way that impedes access to this content, so anything you can grab prior to the switch-over do so.

But before we get started – and absolutely most importantly – it’s crucial to mention how important an SEO’s place at the table is during the site redesign phase.

A side-by-side analysis of what is being proposed is critical to the whole success of the project, and indeed the entire businesses’ marketing strategy.

A website is after all the most important marketing tool that a business has. So take a good look the following typical types of changes that would be detrimental. 

Design & development considerations

1) Renaming URLs will be detrimental to SEO

Adding a directory in the URL path may have an impact on SEO as it is perceived that the content is now one directory further away from the root e.g. /holidays-in-spain/ becomes /holidays/spain/

Changing the name of a file may have an impact on SEO as the primary search term features in the existing URL whether or not anchor text is used in the link or simply the URL itself e.g. /spain-holidays/ becomes /guides/mainland-spain/

2) Changing H1 headings will be detrimental to SEO

e.g. Spain Holidays becomes Mainland Spain

3) Not replicating a special offers page will be detrimental to SEO

If you’ve got one it’s likely to not only be a very useful and therefore bookmarked page, but is also likely to act as an optimised anchor text directory to lots of popular deep content that will need to remain linked to if you want your previous performance to continue.

4) Not using correct H heading mark up (H1, H2, H3, etc.) will be detrimental to SEO

5) Removing the breadcrumb trail will be detrimental to SEO (and user experience)

6) Dropping homepage promotional links is likely to have an impact on SEO

What tabbed sub navigation and offers based quadrants currently feature on your existing homepage?

Don’t ignore any past SEO efforts to get more anchor text rich category links into deep content from your most important piece of real estate.

SEO considerations for site redesign and migration

Please refer to Modi’s original article side-by-side with this one as I have deliberately not duplicated his detail on each of these steps (which are named the same for ease of use).

In order of priority:

First things first

1) Measure rankings

The one thing that you can’t do after the fact – benchmark where your current website ranks before you make any changes.

Ideally you’d have this over a few weeks/months.

During dev:

1) Block crawler access

This should be applied as a universal rule, however you may well want to crawl the dev site using 3rd party crawling tools to pull back various information so you may just want to block search engines in your robots.txt file.

Also add a meta robots noindex to all pages.

2) Crawl the legacy site

Running both Xenu Link Sleuth and Screaming Frog (see the appendix of Modi’s post) will give you a robust picture of the sites’ current performance.

Xenu will show you how many current html pages are published (compare this to how many Google has indexed using the site: search modifier).

It will also create an HTML site map which can be useful for other teams too. Screaming Frog will show you Page Titles, Meta Descriptions (and their lengths), page depth, number of internal links, etc… lots of factors that support the current ranking of the website.

Don’t forget the importance of internal links for SEO. 301’ing external links and then approaching web masters to update them to the new published URLs is one thing, but not carefully analysing your internal links and their anchor text to determine how many you’re going to need on the new site and with what anchor test is a foolish over site.

3) Export top pages

Do this for pages with most incoming links, yes, but also look at the most trafficked pages on the website in Google Analytics as well as Google Webmaster Tools.

What pages are you replacing these with on the new site and how are they likely to perform in comparison?

4) URL redirect mapping

Critical to get this right and to plans (and test) this in the dev environment and the live environment before you go live.

Don’t forget to use the correct type of redirect, and that links passed via a 301 redirect do lose some of their clout, so remember to take the time to update them.

5) Prepare XML sitemap (s)

The quickest way to get the new site indexed will be via an XML sitemap submission to Google & Bing Webmaster Tools.

6) Address duplicate content issues

Hopefully your dev team will come to you asking for solutions to this one rather than you having to spot them and shake the tree, but I somehow imagine that you’ll have to cast a keen eye over what’s being built to ensure that you’re using canonical tags correctly.

On launch:

1. Check 301 redirects
2. Notify Google via Webmaster Tools (migration)
3. Fix broken links <link>
4. Monitor crawl errors

Post-launch:

1) Update most valuable inbound links

Whilst you are in this process don’t forget that this also gives you a chance to request an update to the anchor text of the link.

Be careful though – ideally only one in every four links to a page should have optimised anchor text.

2) Build fresh links

NB If you’ve introduced new category pages and content not featured previously on the old site it’s very important to build new links into these deep content pages.

3) Measure impact/success

Watch the switchover from the old URLs to the new ones in Google in earnest – daily monitoring your site:URL searches; look at real time analytics; set up alerts in GA and GWT; monitor traffic daily compared to previous periods – week, month, year… with any luck you’ll see figures going up!

But most of all keep an eye on your bottom line with page level analysis so you can quickly pull in the relevant teams.

Woe betide any developers who think their job is done after they’ve flicked that big Go Live lever!

Final notes on link equity

At it’s highest level SEO can be seen as on-site optimisation & off-site optimisation. Both factors are important in equal measure.

Optimised content on the site will only be as relevant and visible as the number and quality of the links it has pointing at it.

Links cut through these factors as both on-site and off-site links are important in equal measure. In fact maximising the use of your internal links can have a hugely positive impact on your overall SEO.

To successfully complete a site migration existing URLs (particularly those with links pointing to them) need to be redirected using the correct permanent 301 redirect to their corresponding new addresses. This method minimises traffic loss and is the only way to help search engines correctly update link destinations.

However passing links through a 301 redirect will incur a slight drop in link equity and every effort should be made following go live to manually request external link updates and to correct all internal links to their new addresses.

The value and importance of a link is based on two factors: 1) the strength of the page (and domain) that the link to your site is on, 2) the anchor text contained within the link.

Summary

Get in early. Work with your designers so that they understand enough of the principles to help you improve things for SEO rather than hinder them.

Work with your developers to make sure they don’t cut corners by using frames or hidden content, but most importantly work to provide clarity to your client that whatever you do there is very likely to be fluctuations that you have no control over.

Using this and Modi’s guide and ticking all your check boxes is best practice. Selling in time to build new links and reconfigure old ones currently is a key trick, and don’t forget that the designers bear as much responsibility for any change in performance of the new site as you do.

Here’s Modestos Siotos’ article  on site migration from Moz. 

Posted by Dr-Pete

This post is a part of the “Starting Over” series, the story of starting a blog (MinimalTalent.com) from scratch. See the end of the post for links to the rest of the series.

In parts one and two, I showed how I got my blog off the ground, indexed by Google, and just starting to rank. Now, it’s time to dive in and sand off any rough edges, before they cause future SEO injuries.

(1) Spot-check the SERPs

Marketing automation tools are great, but sometimes we get so enamored with those tools that we forget they only offer a window into the big picture. Early in a site’s life, I’m a big believer in actually typing in searches and seeing how your results look in the wild. The first time I started ranking for the phrase “minimal talent,” it looked something like this:

On the bright side, the site was getting picked up on Google+ (thanks, Jeremy!). Unfortunately, Google was creating a snippet from my first blog post. Why? Well, I hadn’t actually specified a Meta description. Sometimes, even the professionals forget the basics. Once I fixed the problem, I kept watching and eventually saw this:

There’s a wealth of information in this one image. I learned that Google was using my Meta description, but that it might be a bit long (note the odd jump to mid-sentence). I learned that Google was picking my authorship attribution and displaying my profile picture. I learned that my title wasn’t getting cut off. I learned all of this by just opening my eyes and looking.

(2) Google Webmaster Tools

Ok, now that we’ve at least made a few sanity checks with our own eyes, let’s see what the tools have to say. First, is Google indexing the site the way we’d like them to? Since I set up an XML sitemap, I can just go to “Crawl > Sitemaps”, and see something like this:

I’ve submitted 8 pages, and all 8 were indexed – so far, so good. Of course, the “indexed” count on this page only tells you which of the URLs in your sitemaps have been indexed. To get a glimpse at Google’s full index stats for your site, go to “Google Index > Index Status”:

The total count is right in the ballpark of my sitemap count, which, at least in my case, is good. Of course, Google didn’t index any pages before the site existed, so the graph really isn’t that useful. Over time, though, it can show you any unusual trends.

Keep in mind that, for large sites, you can’t expect every single page to be indexed, and that’s often not even desirable. The more you break up your sitemaps, the more you’ll be able to spot problems. If you see your total index count really take off, or you know it’s just way too large (your site has 500 pages, and Google has indexed 25,000), then this could be a sign of runaway URL parameters and duplicate content.

Finally, let’s make sure I don’t have any obvious crawl errors. Go to “Crawl > Crawl Errors” and you should see an overview like this:

I’ve got two “Not found” (404) errors, which really isn’t bad at all. I’m a bit concerned that my initial WordPress “Hello World” post is popping up, so let’s click on that:

The “Error details” aren’t particularly useful here, so I’ll go straight to “Linked from” and can see that the bad URL was on the page itself (a non-issue) and the home-page. Looking at the home-page source code, this link is now gone. So, Google just crawled the site a bit too early, and this problem should take care of itself.

(3) Moz Analytics

While Google Webmaster Tools has a lot of useful information, there can be pitfalls to getting the story from just one point-of-view (especially when it’s Google’s). Let’s look for any crawl issues in Moz Analytics, starting with “Search > Crawl Diagnostics”. Toward the bottom of the page, I get this summary:

Problems are sorted (left-to-right) from high priority to low priority, but my job this time around is pretty easy. I have 38 occurrences of one error, “Missing Meta Description Tag.” This is problematic not just because of the error, but because I really don’t expect to have 38 pages of the site crawled. So. Let’s drill down and look at a few sample pages…

A quick spot-check of the site reveals that these pages do not, in fact, have custom Meta descriptions. While this isn’t mission critical just yet, I should add them soon for my main pages.

As for the 38 crawled pages, it looks as if Moz Analytics is crawling my comment/reply pages. Looking at the source code, these pages have two Meta Robots directives and a rel=canonical tag in place, which is probably giving the crawlers some grief. It’s probably not a big issue, but let’s make sure that Google isn’t indexing these pages, by using the “site:” operator with “inurl:” on the comment/reply URL parameter. Entering the following into Google…

site:minimaltalent.com inurl:replytocom

…results in no documents found. So, at this point, it looks like Moz is being a little overprotective. It may be worth removing either the canonical or Meta Robots down the line, to make sure I’m sending Google clear signals.

Now, let’s look at what really matters – have my rankings improved? Or, at the very least, are they stable?

It’s looking good. I took the top spot for my brand name (“minimal talent”), kept the #1 spot for my tagline, and have even moved into the top 10 for “minimalism 101”. I don’t expect to be ranking for “minimalism” or “yahoo logo” any time soon – these are stretch goals at best. What’s important is to see gradual progress, even if that progress isn’t always as fast as you’d like.

(4) Google Analytics

Are these rankings helps my traffic? Honestly, only a tiny bit. Here’s the graph of sessions for the first couple of months:

It’s not a bad graph, as graphs go, but the spikes correspond with blog posts and almost entirely with traffic from social media (at this point, primarily Twitter). The small increase in traffic between posts toward the right side of the graph is a good sign, and some of that is coming from Google.

I think this graph really illustrates the dilemma of modern SEO. You aren’t going to get search exposure without first building traffic and interest somehow. For me, social is one obvious tool, but for the first few months of a project that means a sustained effort on an established network. For someone with no network at all, the build-up is going to take even longer.

Recapping Parts 1-3

I hope this short series has at least given you some insight into getting started and how the pieces can all come together. I hope it’s also not entirely bad news – ranking in 2014 isn’t easy, but it can be done, and getting the basics right does still matter quite a bit.

We’re going to put this series on hold until something interesting happens to Minimal Talent that’s worth talking about. If anyone has specific questions about getting started or about the site’s successes or failures so far, please chime in.

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don’t have time to hunt down but want to read!