This Month in Search [December 2022]

Well, it’s been a while since I’ve done one of these – I’m sure nothing much has changed… oh.

Monday 12th December

A bit of perspective

Some queries which return featured snippets will now provide several ‘perspectives’ where there may be multiple competing answers. I haven’t been able to replicate the result from a UK mobile device, but it seems like the rollout has occurred in the US. The new feature is covered on SE Roundtable here.

More ChatGPT stuff

I’ve mentioned my feelings on ChatGPT in a recent blog, but there’s a detailed look at the tool over on Search Engine Land which also tackles some of the problems for SEO I touched upon last week.

Friday 9th December

HTML sitemaps are unnecessary and spinners are bad

A post on SE Roundtable covers a statement on Mastodon from Googler John Mueller, who said:

I changed my mind on HTML sitemaps over the years, they should never be needed. Sites small & large should always have a clear navigational structure. If you feel the need for a HTML sitemap, spend the time improving your site’s architecture instead. HTML sitemaps fit into the same bin with HTML loading spinners.

Pretty simply, right? I’m not sure I’d go as far as putting them in the same bin – even Mueller later on stated that spinners (the spinning wheel of death) are actually bad for SEO, while you couldn’t say the same of an HTML sitemap which was historically a net positive for both user and search engine and, at worse, is now probably neutral in usefulness for SEO. Still, I’ve used HTML sitemaps fairly regularly and in the recent past – they’re a great place to find contact details on large sites that clearly don’t want to be contacted, for example.

Content ideas from Search Console

Really interesting experimental feature ‘content ideas’ has been spotted in the wild (covered here by Search Engine Land). Who knows how useful it will be, but it’s always nice to have a new place to find prompts for content.

Thursday 8th December

Programmatic transparency study

It may be no surprise to brands spending money on notoriously opaque programmatic ad buying, but it’s pretty difficult to find out what’s going on with the money once it leaves the account. How much goes to publishers? How much to platforms? A study is currently struggling to make headway with such questions in the US and the issues being faced are covered in an interesting piece published on Digiday.

Who broke the internet?

Is Google getting worse? An SE Roundtable article covers a poll of 342 SEOs that found a split between ‘Google definitely worse: 40.4%’ and ‘Web worse, Google dealing: 41.2%’. I think this misses the important factor that a lot of the content now online was created specifically to make money from ads or direct purchases, is often duplicated or reworked across a number of sites (he says, offering comment on a story appearing elsewhere about a poll made on a social media platform). The question has been floating around the ether for about 12 months (well, for a couple of years, but it came to prominence earlier in 2022), but has been boosted by a recent episode of the Freakonomics podcast which asked the question of Marissa Mayer, 20th Google employee.

As you’d expect, Mayer came down firmly on the side of ‘web is worse’ and, to be honest, I’m inclined to offer qualified agreement – I think Google is definitely better at what it does than 10+ years ago (though the proliferation of ads is reaching beyond saturation), but the web is partially getting worse because of how easy it has become to rank a new drop-shipping or affiliate site. It’s a secondary income for many in the industry (not me, I have neither the time nor motivation), and it’s noticeable how many searches are filled with content generated at scale by scraping or ML programs looking to generate ad revenue. So yes, Google is better, the web is worse, but you can’t say that Google remain blameless in the diminishing quality. There’s a lot of work to do and, in the meantime, there’s Reddit.

Wednesday 7th December

Meta-averse

In a move that might sound familiar to Australian and Canadian news outlets, Facebook/Meta has threatened to de-platform all US publishers if its forced to share some of the money it makes at least in part from being one of the go-to sites for news. They failed to follow through on the threat in the previously mentioned countries, however, so it’s likely another tantrum from a platform that has been flailing in the wake of privacy changes, an aging demographic and corporate scandals. The story is covered in more detail on Search Engine Journal.

Search for the untidy browser

Google has announced an update specifically to shame me for the number of tabs I use and the number of unread articles I’ve bookmarked. In a shockingly personal update, Google has unveiled a series of new ways to search Chrome – @tabs, @bookmarks and @history which will allow you to search through the horrible mess you’ve made of your tabs and bookmarks while facilitating a search for that one tab you actually managed to close six weeks ago and are now regretting.

Tuesday 6th December

Location pages

Interesting piece from Search Engine Land on location pages which goes some way to explaining the de-indexing issues some sites are having with location specific service pages – many of which offer little or nothing to differentiate them from their neighbours and serve solely as doorway pages (pages aiming to serve as a doorway to the rest of a site by targeting a specific keyword). Don’t do it is the simple takeaway (and has been for some time), but interesting to see some evidence of how the Helpful Content updates have impacted them.

Semantic HTML

I’ve covered a lot of this in various blogs here and elsewhere, but this piece by OnCrawl is an excellent resource for anyone wanting to augment their schema efforts with appropriate HTML notation. Not news, but well worth bookmarking.

New Microsoft features

If Microsoft had even half the market share of Google, it’s difficult to see how it could be prevented catching up quickly with advertising spend. The ad offering on Microsoft seems to have been constantly improving over the last few years and integration with Google for imports has also been getting better. Brands with older audiences (the demographic still more likely to Bing their queries) should pay special attention to this article on Search Engine Land.

Helpful Content ranking system goes global

As of the 5th (US), the Helpful Content update, first introduced in August, was updated(?) to “work across content globally in all languages. The rollout may take up to two weeks to complete.“

Monday 5th December

Some Google stories

There’s been a shift to differentiating between specific algorithms and ranking systems over the last month, but one thing has remained consistent – the weekly SE Roundtable blog on an unconfirmed weekend update. The last three or four years have been hugely volatile, but these articles provide a service in demonstrating the continuous nature of ranking fluctuations these days. I’ve said a million times (but am no closer to finding out if I’m right) that this is likely due to the various machine learning algorithms ranking systems currently active in search – there seems to be a process of constant testing which makes for interesting times.

In other Google news, Gary Illyes has offered some advice to owners of JavaScript heavy sites (and it’s not ‘stop having a JavaScript heavy site). The Google Analyst has advised that sites heavy on JS should ensure that content is being served first as long load times for main content are causing pages to be seen as boilerplate duplicates:

Every single SEO has made the same joke today – but it didn’t stop me chuckling to myself – as Google decided to rank everyone on page one of the SERP. Search Engine Journal covers the latest move to unify the mobile and desktop experience with the rollout of continuous scrolling.

Last Google story for the day – again from SE Roundtable is on Google’s confirmation that 4XX status codes (except 429 – too many requests) don’t waste crawl budget. As ever, it’s worth pointing out that you’d need a pretty big site to be impacted by any kind of crawl budget problem, so it’s not something the vast majority of sites need to worry about.