The big search news in March has been all about Google’s planned rollout of the move to mobile-only indexing. But that’s not all that happened this month! We also saw an (as yet unconfirmed) Local search algorithm update, an explanation of Google’s “serving index” from Gary Ilyes, and an announcement that major SEO tool provider SEMRush is going public.
In this roundup you’ll also find us sharing technical tips and tool features to help make your workflow easier, plus our favorite recent insights and research from others in the industry, including an experiment from SearchPilot testing the impact of removing numbers from title tags. And here at Tight Ship HQ we’ve also been discussing Mike King’s prediction that Mobile-Only Indexing will cause “the biggest SEO shake up of 2021”.
Read on for our monthly news and insights roundup for March:
Industry News and Updates
General News
- Google showing fewer featured snippets since February 19th
- Jake: “Looks like this one has bounced back since the article was published on Mar 1st. Maybe a large test on Google’s side?”
- Google talks about its Serving Index and Index “Shards”
- Google’s Mobile-first indexing deadline is around March 2021
- Microsoft Email: Use the Bing WordPress URL Submission Plugin
- Large unconfirmed Google Local Search algorithm update (Mar 14)
- Google PageSpeed Insights now supports HTTP/2 – and scores might go up
- SEMRush to go public with $144M revenue and 67K customers
GSC and BWT Updates
- Google Search Console: Changes to Web Story Performance Report
- Video: Google on Crawl Budget and Search Console Crawl Stats Report
- Ben: “Some items I found helpful from the video:
- Total crawl requests does not count any resources outside of the site (for example, CDN)
- Ben: “Some items I found helpful from the video:
- If Google caches a page resource used by multiple pages it’s requested the first time
- Average response time does not include retrieving page resources such as scripts, images, and other linked/embedded content, does not account for page rendering time
- Crawl Discover – when Googlebot discovers a URL for the first time”
- Google Search Console: new message about “requirements for SharedArrayBuffers”
- Mike: “It seems the required headers are for security, to prevent hack-injected scripts from executing alongside approved scripts, but the SharedArrayBuffers themselves are generally used to improve application performance by allowing for more control over JS execution threads. This seems to have been done to address the Spectre attack method of speculative execution scripts, which Google engineers found and exposed. Here’s a clarifying update from Google.”
Tips, Tricks and Tools
- Crawling Wayback Machine with Screaming Frog (Mike)
- It’s pretty easy to crawl Wayback Machine with Screaming Frog and scrape the first saved date (the first time Wayback Machine stored a snapshot) for a list of URLs. This CSS selector works currently (JS rendering required):
#react-wayback-search > div.captures-range-info > span > a:nth-child(1)
URLs to crawl are just https://web.archive.org/web/*/{pageurl}
(for example: https://web.archive.org/web/*/https://www.domain.com/landing-page)
- What it’s good for:
- This is a good way to figure out when competitors added specific pages (although it will obviously miss URL changes and cases where the page was not crawlable by the Wayback Machine for whatever reason).
- You can also use it to check topic/content of legacy URLs.
- Note that for smaller sites data will be spotty (obscure pages don’t end up in the archive).
- Use Botify to test segmentation against a URL list (Ben):
- What it’s good for:
- This is useful for testing page type rules not available in crawls.
- What it’s good for:
- Using regex with Linux command line to output lines that match to a new file, and prepend the file it was found in (Ben)
- Here’s the an example that we used to identify where hreflang URLs were matching a particular URL structure:
egrep -m 50000 “/pd/[A-Z,0-9]{10}” /drag/and/drop/list/of/file/names > /file/path/and/name/to/output/or/drag/and/drop.csv
- Full example:
egrep -m 50000 “/pd/[A-Z,0-9]{10}” /Volumes/GoogleDrive/My\ Drive/Example\ \|\ Tight\ Ship/SEO\ Dev\ \|\ Example\ \|\ Tight\ Ship/hreflang\ export\ for\ asin-only\ check/export_617497_1048381_example-us_2021-03-17_01-34.csv /Volumes/GoogleDrive/My\ Drive/Example\ \|\ Tight\ Ship/SEO\ Dev\ \|\ Example\ \|\ Tight\ Ship/hreflang\ export\ for\ asin-only\ check/export_617507_1048391_example-ca_2021-03-17_01-59.csv /Volumes/GoogleDrive/My\ Drive/Example\ \|\ Tight\ Ship/SEO\ Dev\ \|\ Example\ \|\ Tight\ Ship/hreflang\ export\ for\ asin-only\ check/export_617509_1048389_example-de_2021-03-17_01-54.csv /Volumes/GoogleDrive/My\ Drive/Example\ \|\ Tight\ Ship/SEO\ Dev\ \|\ Example\ \|\ Tight\ Ship/hreflang\ export\ for\ asin-only\ check/export_613510_1048397_example-fr_2021-03-17_02-19.csv /Volumes/GoogleDrive/My\ Drive/Example\ \|\ Tight\ Ship/SEO\ Dev\ \|\ Example\ \|\ Tight\ Ship/hreflang\ export\ for\ asin-only\ check/export_617505_1048396_example-es_2021-03-17_02-17.csv > /Volumes/GoogleDrive/My\ Drive/Example\ \|\ Tight\ Ship/SEO\ Dev\ \|\ Example\ \|\ Tight\ Ship/hreflang\ export\ for\ asin-only\ check/hreflang_asin_only_pds_example.csv
Insights and Hot Takes
- Tight Ship hot take: “Why listening to Google reps is confusing” (Mike)
- John Mueller recently said there’s no need to worry about bounce rate for SEO. But back in 2017, Google’s head of Google Brain project in Canada said that the algorithm was using return-to-SERP and dwell time – that they were training models on those behavior signals.
- Maybe the algorithm isn’t using your site’s bounce rate / return-to-SERP rate directly, but it stands to reason the things you do to improve your page content and experience (maybe using bounce rate as one measure for the improvements you’re making) can satisfy a model trained on those metrics – thereby improving your rankings. It’s all pretty abstract, but “don’t worry about bounce rate” seems oversimplified.
- John Mueller recently said there’s no need to worry about bounce rate for SEO. But back in 2017, Google’s head of Google Brain project in Canada said that the algorithm was using return-to-SERP and dwell time – that they were training models on those behavior signals.
- SearchPilot: Removing numbers from article title tags results in -16% drop in organic traffic [Experiment]
- Jake: “We actually tested something similar on a real estate client with “top 20 results” type language and saw negative impacts from including that language. Make sense for their industry as users are likely looking for quantity of apartment listings to ensure they are seeing all available.”
- Marshall Simmonds: “YOY Metrics are about to enter the COVID slump”
- Mike King: “Why Mobile-Only Indexing will cause the biggest SEO shake up of 2021“
- This one sparked quite a discussion at Tight Ship HQ (and across the SEO Twitterverse)! The article is arguing that for many websites, there are big differences in both content and links between desktop and mobile homepages, which will have a significant impact when Google completes its shift to mobile-only indexing.