Google’s Helpful Content update is coming

The update targets content designed to attract clicks rather than inform readers.

In the announcement post, Google's Danny Sullivan explains that it will tackle "content that seems to have been primarily created for ranking well in search engines rather than to help or inform people". The aim is to surface a more significant proportion of unique and authentic information rather than content that's aggregating information from elsewhere.

This update is not attacking SEO or designed to kill good SEO practices. But it may well impact SEO techniques considered bad practice or designed to produce "SEO content". For example, there's been growing talk about the use of AI to generate content at scale.

Mark Chalcraft, Head of SEO and Content

As per Google's post on its Search Central site, it is introducing this new site-wide signal to assess how much of a site's content is (considered to be) unhelpful.

Sites where this applies are likely to lose out when the update takes hold, but this does not mean that their most helpful content will not rank if other signals indicate that it should.

The signal is weighted – the more unhelpful content a site has, the heavier the impact will likely be. It is also continuous and automated and uses a machine-learning model. All content on a page will be considered – including user-generated content (UGC).

Google's representatives have stated that the data that powers this update has already been collected, so deleting content now is unlikely to help avoid the consequences (not that some are not trying to do precisely that).

Google also says it will continue to refine the signal over the coming months and will extend it to additional languages in future. It will also "launch further efforts", highlighting the importance of prioritising "people-first content". So, this is the beginning of a new effort, not a one-and-done update.

This update is not attacking SEO or designed to kill good SEO practices. But it may well impact SEO techniques considered bad practice or designed to produce "SEO content". For example, there's been growing talk about the use of AI to generate content at scale.

While using AI to generate content isn't inherently wrong, its use could create the modern, automated equivalent of the content farms that Panda set about to neutralise.

As always, news of a Google update triggers a wave of speculation and rumour, much of which is often unhelpful. This announcement is no different.

Here are some things you may have seen around the web that can't yet be substantiated.

• The impact will be like Panda. The impact of the Panda algorithm in 2011 was significant and wide-reaching. There has been speculation that Helpful Content will have a similar effect. We'll start to find out over the next couple of months, but for now, this is not certain. We can't even be sure whether the impact will be immediate or accumulate gradually over several months.
• Who will be impacted? While Google stated its testing found particular verticals that saw improvements to search results, these are broad and don't tell us much about where to expect the impact to fall. Again, speculation is rife - our advice: wait and see.
• Recovery time. We know that sites will be able to recover, but not how long that can take. Recovery is likely to depend on the effort to eliminate low-quality SEO content.

Still, the information provided by Google suggests recovery may be a slower process "as it determines that the unhelpful content has not returned in the long term".

We've taken a user-first approach to SEO for many years. We're fortunate to have experts in audience research, UX, content design and other fields that work alongside our SEO team.

Our user-first approach includes practices like:

• User testing in various forms to understand how people search, identify behavioural influences and pinpoint things that frustrate or confuse them
• Search intent research which places keywords into the context of the user needs that underpin them. We don't tell clients what keywords to focus on; we show them how to design content to meet the user's needs.
• Search behaviour maps that align keyword usage to stages of the user journey and the moments of truth that deliver successful outcomes
• User-centred design methodology that ensures content production facilitates high-quality user experiences, including consideration of the role of search

This integrated approach means that our SEOs understand the bigger picture. At the same time, we bring a more comprehensive bank of expertise and knowledge to our work through collaboration with a wider team, including agency and client-side.

Don't panic. There's enough in what Google has said to suggest that brands putting appropriate effort into their content are less likely to be impacted than affiliate sites. Connecting consumers with products and services is a core part of what Google Search is for, and that won't change.

• Follow the data. As tempting as it is to rush to judgement, let your data accumulate before making big decisions. Because the signal is site-wide, a methodical approach is essential.
• Prioritise carefully. Your weakest content may act as an anchor weighing down the rest of your site. That means looking away from your main commercial content to those darker, forgotten corners where neglected legacy content lurks.
• Review your governance. Does all your content meet a minimum standard of governance? Do you regularly audit content to identify and update older pages? If all your focus is on producing new content based on trends, you could be storing up trouble for the future.
• Stay in your lane. Produce content on topics you have provable expertise in, and demonstrate those credentials on your site. Content marketing practices have led a lot of brands to publish articles on matters in which they have no expertise. This content isn't helpful for users – people don't go to the bank to ask for advice on car maintenance, so why would we expect them?
• Be evidence-based. What's the rationale for making a change? Is there evidence to support making a particular change? If not, how can you test or prove the idea's validity? "I think…" is a risky way of making a decision, no matter how senior or experienced the person doing the thinking.