Page Rank In The Eyes Of Search Engines

Page Rank

Written by Jeremy Earle, JD

September 8, 2022

This tutorial will go over how search engines determine the importance of signals and how they classify queries.

“How do I rank my page?” is the most common question we ask ourselves as SEO experts.

We should also be asking, “How do search engines rank pages,” if not more so.

Why do Search Engines rank web pages?

Let’s take a step back and consider why search engines rank web pages the way they do before diving into how they do it.

Because it would be less expensive and simpler for them to organize the pages alphabetically, by word count, by date added, or in any other way, they could as well just display them at random.

Of course, they wouldn’t do it. It’s not something you’d do.

In other words, when we inquire for rankings, we must constantly bear in mind that the user we are attempting to gratify is not ours; an engine owns them, and they are being loaned to us.

Because of this, the engine cannot have that user because their ad revenue would decrease if we misuse them.

Think of the situation as a resource page on our site, like this one.

If we recommend a product or service, we do so because we have used it and believe it will benefit our visitors. We will remove them from our site if we learn that they do not.

That’s exactly what the engines are doing.

I have one at my desk and one in my bag when I’m not using it, but for some reason, Google doesn’t pick up messages from any of them.

Because I’ve spent about 20 years watching search engines evolve, reading patents (or more often Bill Slawski’s analysis of patents), and starting each day for many years by reading about the goings-on in the industry from SERP layout changes to acquisitions to algorithm updates, this outline is based on my extensive research.

You should consider the information that I’m about to share as an informed guess that’s hopefully 90 percent accurate.

You may be wondering why I believe 90%, and the answer is that I learned this from Bing’s Frédéric Debut.

There Are Only 5 Steps To Do It!

Ranking a page involves completing five steps.

Not to mention the numerous signal calculations, I’m omitting the more technical issues like load balancing.

Every query has to go through a set of 10 blue links buried in a sea of advertisements to begin its existence as an information request and send it as 10 blue links.

Once you have a firm grasp of how search engines work and who they are intended to serve, you will be in a better position to decide how to rank your pages about those users.

It’s also worth noting that the terminology I’ve used to describe these processes is entirely my creations, not official designations.

Don’t expect any engines to utilize the same vocabulary if you use these terms.

Begin by categorizing

1.In the beginning, we need to categorize the query.

The query’s classification provides the engine with the data it needs to carry out the following processes.

The engines had to use the same signals for all queries before performing complicated classification (read: back when the engines depended on keywords instead of entities).

This is no longer the case, as we will see in the next sections.

A search engine will assign labels such as

  • YMYL
  • Local
  • Unseen
  • Adult
  • Question
  • to a query in the first step of the search process.

However, the search engine would have to first figure out which classes are relevant to a given query.

2. This is the second step. Context

Ranking begins with assigning contextual significance to a given item.

Any relevant information about the user entering the query should be considered by the engine wherever possible.

Even if we don’t inquire, we observe this regularly. They’re all over the place:

We can see them in action right here:


  • Of course, this is an example of a query that I did not enter exactly.
  • Next, the engine must assess the effects of environmental and historical elements on the situation.
  • Since they already know what type of query it is, they can use that information to apply, determine, or pull data relevant to that type of inquiry.
  • As a starting point, below are some examples of environmental and historical data:
  • Several factors to consider include location, time, whether the query is a question, and whether it’s relevant to earlier searches.
  • Is that question anything they’ve ever asked before?

3. Weights are the third step.

Let me ask you something before we begin: Are you tired of hearing about RankBrain?

But don’t worry; it will simply be used as an example in the third phase.

First, an engine needs to determine which signals are most essential in determining page rankings.

When you type in [civil war], you’ll see results like these:

This is a really good result. What if, on the other hand, the importance of novelty has been overstated? We’d wind up with something like:

Freshness, however, cannot be ruled out. I would have given less importance to the show’s authority and more to how recently it had been released. If the question had been, “what are the best Netflix series?”

I’m not interested in reading an article from 2008 that lists the finest DVDs to order through their service yet is heavily linked.

So, with the question type and context components in hand, the engine may now rely on their knowledge of which of their signals apply and with what weightings for the given combinations.

Many talented engineers and computer scientists at Google can do some of this work manually. Still, RankBrain, a machine-learning algorithm designed to adjust signal weights for previously unseen queries but later introduced into Google’s algorithms as a whole, is a system that can handle some of this work automatically.

It is reasonable to conclude that Bing has similar methods because around 90% of its ranking algorithms rely on machine learning.

4. The layout is the fourth step in the process.

Everybody knows about it. In reality, the example of the American Civil War shows this. The layout of the search results page varies depending on the query.

Several different formats can be used for a query depending on what information is being sought and what resources are accessible.

The following is what the first SERP page for the keyword phrase “civil war” looks like:

I’ve developed an educated guess as to how each element is determined.

A moving target requires knowledge of the relationships between things and how those connections are weighted.

We won’t get into it since it’s too complicated for us to cover here.

It is crucial to realize that the many aspects of each particular search results page must be determined on the fly.

After the first three phases have been finished, the engine will look through a database of all the available elements that may be placed on the page and all the possible locations and identify which ones are relevant to the query.

As a side note, the search results pages were produced on the fly previously mentioned.

The engines may store a database of the items they’ve already calculated to meet the most common user intent, although this is more likely to be the case for more common requests.

I believe there is a time restriction after which it refreshes, and I believe it refreshes the entire entry at times of low usage.

The engine can now classify a query, determine the context in which the information is being asked, assign signal weights to that question, and determine the layout most likely to satisfy a query’s multiple intended purposes.

Finally, the rankings have been completed.

5.The ranking is the final step in the process.

Even though it seems simple, it’s not quite as straightforward as it seems.

The 10 blue links come to mind while discussing organic rankings. So, let’s begin there and take a look at how far we’ve come:

  • The user inputs a query.
  • The engine analyses the query type and categorizes it to identify what key criteria apply at a high level based on similar or identical past queries.
  • The engine analyses the query classifications and user-specific signals to determine which signals should have what weights, taking into account the user’s position in place and time.
  • As a result of this information, the engine can also assess which layouts, formats, and supplementary data may satisfy or augment the user’s objective.
  • The engine only needs to crunch the numbers now that it has all this information and an algorithm in place.

To determine the order in which the sites should show in the search results, they will gather the various sites considered for the ranking, apply the weights to their algorithms, and crunch the numbers.

For each element on the page, they must do this in various ways.

All content on a page, including videos, stories, entities, and information, must be sorted by search engines, not just the blue links.


The site’s ranking is simple. The hard part is getting everything in place to make it happen.

You may be wondering how this relates to your SEO efforts. It’s like learning how your computer works from the ground up.

Even though I lack the requisite skills to design a processor, I am well-versed in making a quicker one and how cooling affects its performance.

Because I now know this, my computer runs faster, and I don’t have to update or upgrade as frequently.

The same holds for search engine optimization.

You’ll know your role in the ecosystem if you know the fundamentals of how the engine works.

To put it another way, the engine will be taken into consideration while developing tactics that focus on the user’s needs.

You May Also Like…