The helpful content update is going to cause so much confusion – Marie Haynes

As I write this, we are patiently waiting for the helpful content update to do something. We have heard a few reports of very large and spammy sites possibly being hit, but overall, we have seen nothing. That may change. With the most recent product reviews update, we did not see much movement until the final day of the rollout. It remains to be seen just how strong of an effect the helpful content update will have. I still think it will be huge. 

Google made a big deal of this update. Granted, there have been updates in the past that they got us really excited about that were relatively impactful – Core Web Vitals, mobile friendliness and https sites getting a boost are a few that come to mind. 

I don’t think this will happen with the helpful content update. I really do think it will be big.

We may see a sudden rush of sites getting affected on the last day of the rollout as happened with the latest product reviews update. In that case, the rollout started July 27, 2022, but we didn’t see wild SERP turbulence until after the rollout finished running on August 2.

Semrush sensor showing increased SERP turbulence following the PRU

Or, we may see nothing. But it can still be a massively impactful update if that happens. I think SEOs who help sites deal with traffic drops are going to deal with so much confusion

The helpful content is not a periodic filter like Panda or Penguin

Let’s go back ten years. It’s March 2012. You’re running a site that has top rankings for all of your keywords. You’ve discovered the power of links. In those days, even low quality, self made links were incredibly powerful. A few guest posts, dropping links in comment threads, spinning up nonsensical articles including keyword anchored links to your site…these things didn’t feel good, but they really helped improve rankings.

A site strongly impacted by Google’s original Penguin update rollout

Once the SEO world figured out that cleaning up links was the key to Penguin recovery it sparked a new industry. Prior to 2012 I was a veterinarian with an interest in understanding Google. A few months later, I was spending my days auditing links endlessly, launching me into a new career in helping business owners understand the whims of Google’s algorithms.

With Penguin, we would clean up the problem by removing or eventually disavowing links and then wait till Google decided to run another Penguin update. Panda was the same. If hit, you’d trim out thin content, make content improvements where possible and again, wait.

Eventually, if we did our jobs well, Google would run another Penguin/Panda update. Many sites saw recovery, and new sites were hit.

The helpful content update will act similar in some ways, but there are big differences. 

The helpful content classifier is always running

The helpful content update is always running. Google tells us the new signal they introduced this week is applied by a classifier. This classifier is constantly gathering data to determine if sites have large amounts of unhelpful content. If so, a filter is applied to the site that will suppress its ability to rank. This is a sitewide signal. Sites with lots of unhelpful content will have a bigger suppression than less obvious offenders.

While Panda and Penguin could devastate a site on the day of Panda/Penguin update, the helpful content classifier can hit a site at any time.

It’s always learning

My first thought when I heard the classifier was always running was that this could make for easier decisions when it comes to content cleanup. Let’s say you publish a few thousand pages of mostly autogenerated content and within a few days your entire site takes a hit. Well, now you know! That content was likely deemed unhelpful by Google.

But what if you haven’t published anything new and one day your traffic plummets? Is it the helpful content update? Is it one of the many other things that can cause traffic to drop?

Google tells us the helpful content update is “entirely automated, using a machine learning model”. 

This means that it is going to be continually learning what is and isn’t helpful content. 

Today, it may be that the spammiest of sites that are mass producing thin and perhaps AI generated content are the ones that are hit. But Google’s questions they tell us to ask ourselves go much deeper than just having automated content generation. 

“Does your content leave readers feeling like they need to search again to get better information from other sources?”

“Did you decide to enter some niche topic area without any real expertise, but instead mainly because you thought you’d get search traffic?”

I think that over time, machine learning will help Google get better and better at finding unhelpful content. 

So let’s say that a month from now you wake up, look at your rankings and see that everything has plummeted. Is that the helpful content update? It will be so hard to tell. 

With Panda and Penguin, even if Google didn’t announce an update, we had enough community chatter that we could figure out that Google had pushed out an update. This won’t be the case with the helpful content update because it is always running.

With Panda and Penguin, we had general consensus on what to do to clean up and recover. This will be challenging with a helpful content hit because the algorithm is always learning. What is ignored by the algorithm today may eventually be deemed unhelpful. 

Recoveries will be sporadic

Google says recovery is possible:

“A natural question some will have is how long will it take for a site to do better, if it removes unhelpful content? Sites identified by this update may find the signal applied to them over a period of months. Our classifier for this update runs continuously, allowing it to monitor newly-launched sites and existing ones. As it determines that the unhelpful content has not returned in the long-term, the classification will no longer apply.”

With Panda and Penguin, Google would run an update and on that day many sites that had been working on improving would see recoveries. The helpful content update doesn’t work this way. 

As the classifier assesses your site, it needs to see you’ve got good helpful content and that you’ve maintained that status for a few months. And then one day magically the filter will be lifted (or perhaps reduced). 

Let’s say your traffic takes a hit in September. You won’t know whether it’s the helpful content update that hit you, or something else. You think it was the helpful content update, so you work on content cleanup. And then wait. A few months pass and nothing happens. Should you remove more? Do you need to work harder to make your content helpful? Perhaps something else was affecting your site? Decisions are going to be very difficult.

And so we wait

Before this update launched, Google reached out to several SEOs, myself included to inform us. I have been studying Google updates for over ten years now and this has never happened. After speaking with Danny Sullivan I really got the impression the helpful content update would have a massive impact.

I still think it will.

(Note: This originally started out as a piece for premium subscribers of my newsletter where I postulate on what I think Google is doing. It got really big, so I made it a blog post. )