Google moving toward a continuous “realtime” algorithm?

Even if you have only been doing SEO for a while, you are likely to be familiar with the Google algorithm update. Google updates its algorithm every now and again, regardless of whether it’s something we like or not. This can affect our rankings. The key phrase is “presses button.”

What if Google were to push a button, and the algorithm updated rankings “real-time”, instead of us having to press a button? How would this world look and who would it benefit from?

What does continuous real-time algorithm update mean?

Technology is always changing, but it is clear that Google’s algorithm is also evolving. The search engine will be able to better understand content and evaluate websites as the technology improves. This technology must be integrated into the algorithm. Google must “make these advances a part of its algorithm” as new technology is made available to Google.

MUM is an example. Google is using MUM elements in its algorithm. MUM is still not fully implemented at the time of writing. Based on Google’s announcements, MUM will almost certainly be applied to other algorithmic tasks as time passes.

Google will most likely want to reevaluate its rankings once it has introduced new technology or refined its existing capabilities. Google would want to use these abilities to improve its rankings if it is more adept at understanding site content and assessing site quality. Google then “presses” the button and issues an algorithm update.

Let’s say that one of Google’s machine-learning properties has changed. It has taken in the input over time, and it has been refined. Google could decide to “reintroduce” this machine learning property into its algorithm and reassess pages being ranked accordingly.

These updates are precise and purposeful. Google is “pushing” the button. This is evident when Google announces a core update, product review update, or spam update.

Perhaps nothing says it better than what Google has to say about its spam updates :

span style=”font weight: 400 While Google’s automatic systems for detecting search spam are always in operation, we sometimes make significant improvements to their functionality …. We make improvements to the system from time to time to improve its ability to detect spam and ensure that it captures new types of spam.

Google, in other words, was able create an improvement to a machine learning property and released an upgrade so that it could be applied to ranking pages.

What would continuous, real-time updates look like if this is a “manual” process? Let’s look at Google’s Product Review updates. Google’s Product Review Updates were first released in April 2021. They aim to eliminate product review pages that lack substance, are unhelpful and, if we are going to call it a spade, exist primarily to generate affiliate revenue.

Google uses machine learning to achieve this goal, focusing on specific criteria. Each update, such as the one in December 2021 and March 2022, has a different result. These machine learning devices have the ability to calibrate and improve. They can become more efficient over time as they “learn” – which is a good thing when it comes down to machine learning.

My theory is that rank fluctuates as machine learning properties improve. Google allows machine-learning properties to “recalibrate” their rankings. Google reviews the changes and decides whether they are in its favor.

This process may be known as unconfirmed algorithms updates. However, I’m not saying all unconfirmed update are such. This is why I believe there’s a strong tendency to rank reversals between official algorithm updates.

It is quite common for the SERP to experience a noticeable rise in rank fluctuations, which can have an impact on a page’s ranking. However, those rankings will reverse back to their original position with each wave of rank fluctuations (either a few days or weeks later). This process can be repeated multiple times. This results in a page experiencing rank changes, followed by reversals, or a series reversals.

A series rank reversals that affect almost all pages ranking between positions 5 and 20, which align with across the board heightened rank fluctuations

As I see it, this trend is Google allowing its machine-learning properties to evolve, recalibrate or whatever you want to call it in real time. This means that no one at Google is pressing a button, but the algorithm adjusts to the continuous “real time” recalibrations of its machine learning properties.

This dynamic is what I mean when I ask if we are moving towards “continuous” or “real-time” algorithmic rank adjustments.

What does a continuous real time google algorithm look like?

What’s the point? What if Google adopts a continuous real time model? What would the practical consequences be?

It would make rank volatility far less of a constant. This would mean that ranking volatility wouldn’t be as much of a constant. Instead of waiting for Google, it would be the norm. The algorithm would constantly evaluate pages/sites and make adjustments to rank in real-time.

Another reason would be the lack of need to wait until the next update is released for restoration. Although it is not an exact rule, rank restoration will not occur until the release the next Google update. Your pages will then be evaluated. Pages are evaluated in real time, just like links with Penguin 4.0, which was released in 2016. This would represent a significant change in the “SERP ecosystem” currently in place.

Even more, I believe we have a continuous, “real-time”, algorithm. It is a fact that at least a portion of our real-time Google algorithm exists. Google released Penguin 4.0 in 2016, which eliminated the need for us to wait for a new version. This specific algorithm evaluates pages continuously.

But, outside of Penguin what does it mean to say that we have a continuous real time algorithm?

The case of real-time algorithm changes

The ecosystem’s constant rank adjustments are so important that they have reshaped the volatility landscape.

According to Semrush data, there was an increase of 58% in days that reflected high rank volatility in 2021 compared with 2020. The number of days that showed either very high or high levels of rank volatility also increased by 59%.

The increase in instances that show high levels of rank volatility can be summarized as follows: These trends were examined and I concluded that the cause is the rank reversals. This means that a large amount of rank volatility is due to what I believe to machine learning continuously recalibrating in “real time,” which produces unprecedented levels of rank reverses.

This is supported by the fact that, despite the increase in instances of rank volatility, we didn’t see an increase in the severity of rank movement. This means that there are more instances rank volatility, but not more volatility.

Comparatively to 2020, 2021 saw a decrease of how dramatic the average rank movement was.

Why? This is due to the “real-time” effect of machine learning properties on rankings. We are starting to see micro-movements that correspond with the natural evolution Google’s machine learning properties.

You won’t see huge swings in rankings if a machine-learning property is improved as it learns. Instead, you’ll see a refinement of the rankings that is in line with the machine learning.

As a result, we see a lot more consistent but not as dramatic rank movement.

Final step to continuous, real-time algorithm update

Although much of the ranking movement is ongoing and isn’t dependent on algorithmic refreshes in particular, we are not yet there. Many rank fluctuations are caused by reversing rank positions, as I have already mentioned. These ranking changes are not usually solidified until an official Google update is released, most often an update to the core algorithm.

We don’t yet have a continuous, real-time or continuous Google algorithm until the long-lasting ranking patterns can be established without having to “press that button”.

But, I wonder if this is a trend. Google’s Helpful Content Updates (HCU), for instance, works in real-time.

Per Google :

“Our classifier for this upgrade runs continuously, allowing them to monitor both newly-launched websites and those that are already in use. If it finds that the unhelpful content is not returning in the long-term the classification will cease .”

How can this be? Google allows its machine learning, just like we have been saying here, to have the autonomy that it needs to be “real time” or “continuous”, as Google calls them.

The classifier process is completely automated using a machine learning model.”

Continuous does not necessarily mean constant. The HCU has a validation period that allows for logical verification before restoration. This may also apply if we ever see a continuous real-time continuous algorithm. If we ever do see a real-time algorithm, I won’t be surprised if there is a ranking response after you make any page changes.

The “traditional” button-pushed algorithm update has been less effective over time. A study I did in 2021 revealed that the core updates released since 2018’s Medic Update were significantly less effective.

Data shows that Google’s core updates have less rank volatility as time passes

This trend has continued over time. My September 2022 Core update analysis revealed a significant drop in volatility relative to the May 2020 Core Update.

Rank volatility changes were much less dramatic during September 2022 CoreUpdate than the May 2022 CoreUpdate .

It is a double convergence. Google’s core updates seem less impactful than others (obviously, sites can be slammed equally hard), while its most recent update (the HCU), is constant.

It all points to Google wanting to change the algorithm update release process. Further evidence could be found in the way official updates have been released. The data from the different outlets that cover these updates will show that roll-out is slower, has less volatility, and has a smaller overall impact.

Google would like to move to continuous real-time models, but why?

Why is a continuous, real-time Google algorithm beneficial

What about a continuous, real-time algorithm? Google would love that. It’s quite simple, I believe. Google is winning by having a continuous update that refreshes rankings to reward sites and pages that are relevant (I don’t mean immediate content revisions or optimizations resulting in an instant rank change span>

Which is better for Google’s users and which is less? Continuous-like updates of the best results, or periodic updates that can last months to present new information?

Google is better at updating and analyzing data in real-time. It helps users who are looking for the best results to have rankings that change with every new algorithm update.

It would be much better for users if sites, after seeing their rankings fall, made improvements that resulted into great content and users could access it on SERPs far sooner than waiting months.

Google is able to deliver better content to users faster through continuous algorithmic implementation.

It is also more beneficial for websites. Is it really worth the effort to implement a change to address a ranking loss, then waiting for months to restore your rankings?

Google’s decision to heavily rely upon machine learning and trust its adjustments is a sign that Google is confident in its ability understand content, relevance, authority, etc. This is what SEOs and site owners need. This means Google could be less dependent on secondary signals and more on the primary commodity, content, and its trustworthiness.

Google’s ability to assess pages and content is a good thing for the web. This opens up the possibility of niche sites, and sites that don’t have super-authorities (think Amazons and WebMDs around the globe).

Google’s greater understanding of content leads to more parity. Google’s move towards a more realistic model would be an example of this better understanding.

A new way to think about Google updates

Continuous real-time updates would fundamentally change how we think about Google updates. This would render tracking updates, as we know them, obsolete to some extent. It would alter the way we view SEO weather tools. Instead of focusing on specific times of higher rank volatility, we’d instead pay more attention overall trends over a longer period of time.

Based on the ranking trends that we discussed, I would argue that that time has come. We live in an environment in which rankings fluctuate more than ever before. This has changed the meaning of stable rankings in many cases.

To summarize and simplify, we are moving closer to a continuous real time algorithm for ranking organically in Google’s SERP.

Wix’s Head of SEO Branding is Mordy Oberstein. Mordy Oberstein can be found on Twitter at @MordyOberstein

Sign up for the Search Engine Watch newsletter to get insights into SEO, the search landscape and digital marketing.

Join us on LinkedIn or Twitter to continue the conversation.

The post Google is moving towards a continuous, “real-time” algorithm Search Engine Watch first published this article.

Did you miss our previous article…
https://onlinemarketingagencies.net/?p=6046