Server error! Check the console for details :(

AM - Keyword Tuning

Task is to review and audit the site using the below tools and outline. Once a site is at the point where it's running smoothly, it should only take 10-15 minutes to review everything. For sites that have a lot of problems, 45-60 minutes (or more) is not an unrealistic expectation. This will take especially long on sites that have lots of Crawl Errors from internal pages on the site as there may be a LOT of pages that need troubleshooting and fixing.

NOTE: Ideal is to do Every Step, Every Time
(yes, this time too)
BUT: for the time being do not spend more than 90 minutes on any individual Keyword Tuning task.

Step 1

Check the /blog.aspx page to make sure there's no commented out content on the page. Commented content is still indexed by webcrawlers. On the blog.aspx page, this pollutes the index of EVERY SINGLE blog article.

Step 2

Review the site on Google's Webmaster Tools

• https://www.google.com/webmasters/
• ga@atomic8ball.com account

Make sure the entry you're selecting matches the protocol of the site.

• e.g. If the site has an SSL and is https://www.domain.com/, but the entry in Webmaster Tools is for http://www.domain.com/ then the entry is not giving you accurate information. You need to use the https one and, if there isn't one there, add it.

2.1 Old Search Console

Gears drop-down in top right corner

• we use a standard of www for all site domains, so also make sure you're selecting the entry that includes the www.
• should be on second entry "display URLs as www.domainname.com".
• may require adding an additional verification property.

Search Appearance - HTML Improvements

• meta descriptions
• title tags
• everything should be 0
• if it's not 0, it will give an error such as:
  "Duplicate meta descriptions (pages) 28"
• clicking on the error will show you every page that has the error, all of these need to be fixed, but can be done during Step 3 below.
• strive to fix the duplicates, but be aware that the blog category pages are likely to throw duplicate descriptions when the longest p-tag on the aggregate page matches the same p-tag on the individual post. Can't fix this one.

Search Traffic - International Targeting

• Country
  [x] Target users in: United States
• if it's not check-boxed, please update it to be so.

Security Issues

• should be no issues. If there is a security issue, there's a problem. I haven't seen any show up on any site yet, so will likely need to make an individual review of any issue that comes up.

2.2 New Search Console

Performance > Queries

• give a cursory review of what search items are generating Clicks. Note anything that doesn't involve the company name and seems like a legitimate search query. When reviewing keywords we'll be leaning towards using the keywords that are already shown to be working to generate traffic to the site.

Index - Coverage

Primarily focused on errors, but the "Valid" number should be populated with a good number of items. If you're seeing a "0" in the "Valid" section then there's something wrong. You're probably looking at the HTTP version of an HTTPS site, for example.

So review anything that shows up here and strive to resolve the issue(s). For anything on this section, you can get more information via:

• inspect the URL in WMT by hovering over the URL and clicking on the magnifying glass
• click on "Test Live URL"
• if the page works and matches the then click "Request Indexing"

Some examples:

• Submitted URL seems to be a Soft 404
  • does the page exist?
  • is the page loading content via an iframe and the iframe source does not work? Attempt to fix the iframe, but it may be non-functional now and just need to be deleted.
  • is it a PDF? if so, does it load locally for you? If it doesn't, check back through SVN for an older version of the PDF and replace it. If neither of those work, delete it and the reference, and send an email to updates with the details about where it was, how it was described, and what the file was named.
  • if it's a PDF, does the URL to the PDF that WebmasterTools is referencing have spaces in it? In general, only items that are going through the pr..js user control (Press Releases, primarily) should have spaces in them. Everything else should have the spaces replaced with hyphens.
• Response code 502
  • "Googlebot couldn't access the contents of this URL because the server had an internal error when trying to process the request."
  • Probably some sort of read error, so simply mark it as fixed.
• Submitted URL marked 'noindex'
  • change in prior handling, pages flagged 'noindex' should also be accompanied with a redirect.
  • should this page actually be indexed? (potential copy/paste error on a developer side)
  • all aspx pages and all pdfs that exist on the site get submitted in the sitemap.xml file, regardless of whether they're navigable or not.
  • these should be rare, remove the 'noindex' entry from /thanks.aspx pages and other ones that don't match the rare circumstances outlined further down on the page.

Index - Sitemaps

• there should be a /sitemap.xml with a status of success. If there isn't, submit one.
• if the sitemap is pulling through .node.a8b.co and throwing errors as a result, then CloudFront needs "Cash Based on Selected Request Headers" > Whitelist, with Host added to Whitelist Headers.

Enhancements - Mobile Usability

The most commonly seen issues will be:

• text too small to read
• viewport not set
• clickable elements too close together
• content wider than screen

If these occur on .html files contained within an /ecampaign/ folder than this is fine, these pages aren't created with mobile in mind as they're targeted at email delivery. If the error is showing on an .aspx file, then check:

• is the site mobilized? Check by…
  • pulling up the page in Chrome
  • hitting F12 to pull up the developer tools
  • hit control + shift + m or click the icon in the upper left hand corner that switches it to mobile view
  • change the width to 425px (you can click on the gray rectangles at the top for pre-determined sizes, one of them is 425px)
  • hit F5 to reload the page
  • does the page look bad? can you scroll horizontally? (the latter should result in a "content wider than screen" error)
  • if the site is mobilized, the problem may be limited to that page. if it's not mobilized (you'll see the entire site just like the desktop except really small) then send an email to updates@atomic8ball.com

Step 3

We'll be reviewing, and potentially updating, the title-tags on every page of the site.

<meta name="description" content="content" />

• goes in the cphHead
• this entry is primarily necessary if the page has no actual text content on it. The system automatically generates this text from the longest <p> tag on the page, so if the page has no <p> tag you'll need to write something for here. Some contact pages have only a form and some phone numbers, for example. Better than writing a meta description, would be to write an appropriate <p> tag entry to introduce the page instead.
• in some cases, the longest <p> tag on the page doesn't contain relevant information for the description, or for stylistic reasons the introductory paragraph has been broken up in to multiple <p> tags (e.g. to enhance readability) and so it may periodically make sense to specify the description for these reasons.
• since it is generated from the longest <p> tag on the page, you may see a duplicate meta description error on pages that have little or no content, and the meta description is being generated from a <p> tag in the footer.
• should be at least 2-4 sentences long.
• does not have a character limit per se, Google has their own algorithm for how much is displayed, and it may be as much as 400 characters. Anything beyond the point that Google opts to display will simply be truncated.
• (old information that is now obsolete targeted meta descriptions at 150-1555 characters or less, this can be ignored now.)

Title Tags

You don't need to do title tags on .ascx, .html, etc, only on .aspx, and they appear in the first line of a .aspx page like so:
<%@ Page Title=""

The title tags are extremely important for indexing purposes, and have a lot of rules and guidelines for how to do them. Once a site has been done properly, this portion of the task is mostly just skimming through the pages to see if anything looks wrong at a quick glance. e.g. someone built a new page and didn't do them properly.

Top-level pages, i.e. the home page and anything that's one step away from it, should get a little extra scrutiny to make sure that the title tags on the page match the content.

Title Tags - Rules

• HARD limit of 70 characters, 71 is too many. Entry should be as close as possible to this limit without going over.
• all words in the title tag MUST appear on the page in some fashion (includes header and footer), otherwise the keyword has lesser/no value.
• no word should be repeated, you're wasting your 70 characters by repeating words.
• special characters should not be used, "&" for example, must be encoded to "&" which is longer than using "and" since it's 5 characters now instead of 3. You can use any alpha (a-z) or numeric (1-9) character, as well as ,.<>/-.
• don't use random separator characters like ::, |, -, etc, they just pad space and consume your 70 digit budget while adding no value. " | " takes up 3 characters, for example, while ", " only takes up 2.
• if you're going to use quotes for some reason, the ' or " used can NOT match the marks used to surround the tag in the first place.
  e.g. Title="this is a 'title' tag" OR Title='this is a "title" tag'

Title Tags - Guidelines

• 1234567890123456789012345678901234567890123456789012345678901234567890
• that is 70 characters, you can put this into a notepad file and make sure you don't go past the limit. You can also paste this into a title tag on the page and adjust the width of your IDE to make it so that 1 character too many will wrap onto the next line.
• the company name is NOT mandatory in the title tag, if you need to fill space at the end you can add it there, but it should generally not be the beginning of the title tag.
• the title tag should strive to be at least sort of conversational English, instead of just keyword stuffing. Minor use of connecting words is allowed "and" "at" "a" "of" "for" etc. You can also use commas to separate words and phrases to make more sense.
• all words should be capitalized except for the connecting words (of, a, the, for, to, etc).
• header tags are good candidates for pulling keywords from, but may need a reality check.
• the keywords you noted from Search Analytics above are good candidates for including on pages where they make sense.
• "about" pages are a good candidate for including the owner's name, as it won't generally have many opportunities to appear on the page. The phone number of the company is also a good candidate on the "contact" page, but can be used as filler where needed.
• try to imagine what someone who needs to land on this page might be searching for. If it's a product and there's a model number, include at least a beginning portion of that.
• can cut out "s" on the end of plural words to save characters, if needed.

Excluding from Indexing

A page that shouldn't be indexed needs a meta tag in the cphHead (or on a submaster page) declaring it off-limits. This won't prevent it from being added to the sitemap.xml file though.

<meta name="robots" content="noindex">

When pages should be excluded:

• the rare circumstance that we want to keep the page in existence with content on it, but we don't want it to be considered 'active'. An example of this might be a career opportunity page that is currently not seeking applicants. This 'noindex' should also be accompanied with a redirect either in the 404..js file (preferred) or in the cphHead.

Iframes

All <iframe> elements should have a title tag on them that gives a synopsis of what the iframe is showing. Besides telling the crawlers what third-party content is being loaded, this is also for disabled accessibility purposes as screen readers for the blind do not scrape any content in iframes.

• title="Google Map location for 123 Main St, Anytown, USA, 99999"
• title="Video - Training Material for Nifty Phone 9000 Series

Alt Tags

Slightly beyond the scope of this task, but adding for completeness.

• alt tags on images should be properly cased, as this is the text that will show when an image isn't displayed.
• alt tags should be conversational, as this is what will be read to a person using a screen reader who is unable to view the image.
• should not say "image of" or "picture of", etc, as it is alt text it is by definition describing an image.
• the alt text should describe the image well enough that you could pick which image is being referenced out of a display of other images.

For example:

• Bad - Network technology
• Bad - man at desk
• Bad - 48194313_xxl.jpg
• Good - Man in white shirt sitting and looking down at his desk with sunlight coming through the window behind him

Which description would allow you to select a specific image out of that lineup?

Step 4

• In the old version, Messages -> delete all the messages

Turn in the task

• when turning in the task, turn it in as "TASKID# KEYWORD TUNING"
• this will make it easier on you or the next person who does this task on this site, as they can look in the SVN Log and see what pages were updated since the last time the task was done. These pages that were updated in the interim likely need a bit more attention than the other ones that haven't been touched.
• if you ran out of time or otherwise didn't get to some aspect of the task, note that in the SVN commit as well. e.g.
  • 12345 KEYWORD TUNING - spent entire task dealing with crawl errors, did not review title tags or meta descriptions