3 New Metrics for Today’s SEO

Longstanding search-engine optimization metrics are becoming less informative. Many do not translate into traffic, owing to Google’s new features in search result pages, such as “People also ask,” local packs, and image packs. Moreover, AI-generated answers, called “snapshots,” will soon appear in Google’s Search Generative Experience.

Even domain authority is outdated. It attempts to assess site strength solely from backlinks despite Google de-emphasizing their importance.

So how do we evaluate modern SEO performance? Here are three of my new go-to metrics.

Brand Authority from Moz

Moz’s new Brand Authority reflects a brand’s search engine strength on a scale of 1 to 100. The metric is mainly based on the variety and volume of branded search queries, including the business and product names.

Moz’s new “Brand Authority” reflects a brand’s search engine strength on a scale of 1 to 100. Click image to enlarge.

Brand Authority helps evaluate and compare a brand’s strength to competitors’. It’s handy for diagnosing ranking losses and gauging viral marketing and public relations campaigns.

There is no “good” Brand Authority — some niches have limited opportunities for exposure. Use the metric to compare sites and tactics. For example, a competitor gaining in rankings could have an elevated Brand Authority score that merits study.

Authority Score from Semrush

Authority Score by Semrush measures the reputation and trustworthiness of a website based on several factors:

• The number and quality of external backlinks. Links from high Authority Score sites increase the grade.
• How the site ranks for relevant, highly competitive keywords.
• Red flags, such as referring domains from the same IP network and a too-high percentage of dofollow links.

Semrush’s algorithm combines those assessments and more to assign an Authority Score from 1 to 100.

It’s a relative metric for comparing domains in similar niches, not an absolute mark. Thus, like Moz’s Brand Authority, there’s no good or bad Authority Score. Use it for higher-listing sites to evaluate the ranking difficulty of specific keywords.

Semrush’s Authority Score combines multiple factors. It’s a relative metric for comparing domains in similar niches. Click image to enlarge.

Visual Rank from DemandSphere

Gone are the days when Google served 10 blue links per page that matched a search query. We could actually estimate organic click-throughs based on a position!

No more. Search results now include visual and interactive elements that engage users, distract them from clicking, and encourage them to scroll deeper. Hence we cannot evaluate SEO performance from rankings alone.

Visual Rank from DemandSphere reflects “share of voice,” the pixel percentage of a search-result page occupied by a site or brand.

The tool analyzes a given query’s top 20 SERP features  — organic listings, ads, knowledge panels, images, videos, and more. It then assigns values to the features based on their prominence and visibility to searchers. For instance, top organic listings likely have higher values than sidebar ads, but lower listings could have less.

Google itself could have a high Visual Rank given the SERP sections with no external organic links, such as answers and ads.

Visual Rank from DemandSphere reflects “share of voice” (SoV), the pixel percentage of a search-result page occupied by a site or brand. Click image to enlarge.

Go (Golang): The Unsung Hero In The SEO’s Toolkit

Go is a young language compared to a language like Python, born in 2007.

It has incredible features such as fast execution speed compared to interpreted languages.

It excels in handling concurrency, making it one of the best choices among other languages.

It is easy to scale and has a great community supporting it.

As a technical SEO/SEO developer who develops various tools for automating the SEO process, one of the key points is being fast and efficient in developing and shipping code without complexity.

In these situations, Go (Golang) shines as bright as the sun.

SEO Automation

When it comes to SEO automation, there are many ways to structure the whole project.

You can write your entire app in a single language and package, or you can split your app into microservices/serverless functions and write each service using the language that performs best for it.

If you are working on a simple automation task, this isn’t your concern.

However, as the project grows, you may encounter productivity bottlenecks as an SEO developer.

Who Is An SEO Developer?

When you search to know more about SEO developers, you’ll find some sources that refer to SEO developers as front-end engineers who know about SEO.

However, we know that in an ideal website structure, the app front-end only serves as a shell for the app, not the data.

A back-end engineer is responsible for retrieving the data and returning it to the app’s front end.

The front-end is responsible for creating components with SEO in mind, such as headings, setting titles, meta descriptions, etc.

Furthermore, there are technical SEO experts who also code.

With all these statements in mind, some of us are SEO developers.

Go, JavaScript, and Python: Choose The Right Hammer Based On The Nail

As SEO pros, most of us use Python as the primary language for coding.

With Python, we can perform data analysis on our data, create charts, code web applications using frameworks like Flask, and instantly get a package that crawls the entire site for us, etc. Python is our Swiss Army knife that allows us to do anything.

But what if a new language could help us improve our productivity in some of these tasks?

There are different ways to compare these languages, including their type system, how they handle concurrency, etc.

Before anything else, I assume you are an SEO developer who knows Python/JS.

If you are new to coding as an SEO, Python is the better choice for you in most cases, rather than learning JS or Go.

Below is my preference for the task: each language can better serve me as an SEO professional:

• Python: Analyzing data and performing ML-related tasks.
• JavaScript: Writing Apps Script, working with Google Tag Manager as an advanced user. Interested in front-end development.
• Go (Golang): Developing web applications. Interested in back-end development.

Go Facts

Let’s become more familiar with the Go language.

Go History

If you are reading this section, you’ve decided to learn Go as your new coding language.

Let’s briefly look at Go’s history.

Go was developed by Robert Griesemer, Rob Pike, and Ken Thompson at Google in 2007 and launched as an open-source programming language in 2009.

It’s interesting to know that it was inspired by Python (created in the late 1980s) for its simplicity.

Go Is A Compiled Language

Besides the fact that compiled languages are faster in execution than interpreted languages, when I code in Python and want to execute it on another machine, I find myself struggling to obtain the packages, resolve package conflicts, and so on.

However, with Go, I can easily build the code on my non-Linux machine for the Linux machine where I intend to run the code.

Then, I use the “scp” command-line tool to copy the executable to the Linux machine.

Go Has A Fabulous Standard Library

One of the amazing aspects of Go is that you can explore its standard library.

If you want to make a request, instead of downloading the “requests” package in Python, you can use the built-in package.

Similarly, if you want to create a web server, you can do it without installing any additional packages. In many cases, you only need to consult the Go standard library to find a solution for your problem.

Go Is Fast

When we say Go is fast, we can assess its speed from various perspectives. Here are some of these aspects:

• Because of its simplicity, you can have a nice and fast development experience.
• It’s an effective garbage collection that manages the memory for you.
• Concurrency is one of the things that Go is famous for, and it’s easy to set up (unlike Python).
• Since Go is a compiled language, you get relatively faster code execution compared to interpreted languages.

What Are The Tools For Coding Go?

There are several options for coding in Go. Below, you can see a list of these tools:

• Visual Studio Code (VS Code) – Free.
• GoLand – Paid.
• Vim/Neovim – Free.

Personally, I code in GoLand, but with the IdeaVim plugin that brings Vim motions to my IDE and makes my life easier.

If you want to use VS Code, install the official Go plugin for it. If you prefer Vim/Neovim, don’t forget to install Go LSP.

How To Install Go

You can easily follow the instructions the Go website provides to install Go based on your operating system.

How To Write Hello World In Go

Let’s GO.

After installing the Go language on your computer and verifying that it’s installed, create a folder wherever you want and name it “hello-go.”

Then, on your terminal or Windows Subsystem for Linux, use the “cd” command to navigate to the folder you created.

Since you have installed Go, you can now access the Go command line on your machine.

In your command line, execute the “go mod init hello” command.

This command will create a “go.mod” file, which declares the module name, the required Go version, package dependencies, and more.

If you don’t understand it yet, don’t worry – take the time to learn it without stopping at this moment. If you’re familiar with Poetry for your Python projects, they have some similarities.

Now let’s create a file to write our code in and name it “hello.go.” If you are creating the file using your terminal, you can run the command “touch hello.go“.

Let’s open our text editor/IDE to write our first Go code. Below, you can see the code, and then I will explain it to you.

package main
import "fmt"
func main() {
fmt.Println("Hello GO!")
}

There Are Different Elements To Consider About The Above Code

• Package name: We named it “main” to indicate that this is our main package in Go.
• Import statement: Similar to Python, we imported the “fmt” package from the standard library, which is used for formatted I/O.
• Main function: This serves as the entry point for Go to run our program.

Maybe now you are asking yourself, “Alireza said Go is easy, and what are the hacks above? Python is simpler, etc.”

There are typically some differences between Python and Go, but at this moment, let’s consider that we write our functions outside the main function and then call them inside the main function.

When Go wants to run our program, it knows it must run the main function. So, if there is a call to another function, it goes and runs it.

If you are familiar with languages like C, the above code will be much more familiar to you.

How To Run/Build Our Program In Go

After coding our program, we need to run it to see “Hello GO!” as the output. To do this, we run the command “go run hello.go” in the command line.

As you can see, we have our output.

As I mentioned, Go compiles our code and is not similar to Python, which uses an interpreter. Therefore, we can obtain an executable file and run it!

When we use the “go run” command, it creates the executable file on the fly and runs it without saving it. However, by executing the “go build hello.go” command, we obtain our executable file as output with the same name as the file we passed to “go build.”

After running “go build hello.go” we should have a “hello” file as the output, which is an executable file. We can run it from the terminal using the “./hello” command.

Python Code Equivalent In Go Code

Now that we know the basics let’s look at creating a variable, a for loop, making an HTTP request, etc.

Below, I will write both Python code (which I assume you are familiar with) and the equivalent code in Go for a better understanding.

The project “Golang for Node.js Developers” inspired me to write in this way. I also created a project called “Go for Python Developers” with as many examples as possible for me, so don’t miss it.

Let’s GO!

Variables

Python:

mutable_variable = 2
CONST_VARIABLE = 3.14 # There isn't a way to define a constant variable in Python
a, b = 1, "one" # Declaring two mutable variables at once

Go:

var mutableVariable int = 2
const ConstVariable float = 3.14
var a, b = 1, "one" // Go automatically assigns types to each variable (type inferred), so you can't change them later.

Besides that, the above example shows you how you can create variables in Go.

You can also see the different ways of naming variables in Go and Python, as well as the various ways of leaving a comment.

Data Types

Python:

string_var = "Hello Python!"
integer_var = 2
float_var = 3.14
boolean_var = True

Go:

var string_var string = "Hello Go!"
var integer_var int = 2
var float_var float = 3.14
var boolean_var bool = true

For loops

Python:

i = 10
for i in range(10):
print(i)

Go:

// Initial; condition; after loop
for i := 0; i < 10; i++ { // Using the shorthand syntax of declaring a variable in Go (mutableVar := the value)
fmt.Println(i)
}

While loop

Python:

counter = 0
while counter < 5:
print(counter)
counter += 1

Go:

var counter int = 0
for counter < 5 {
fmt.Println(counter)
counter += 1
}

If/Else:

Python

age = 25if age >= 13 and age <= 19:
print("Teenager")
elif age >= 20 and age <= 29:
print("Young adult")
elif age >= 30 and age <= 39:
print("Adult")
else:
print("Other")

Go:

var age int = 25
if age >= 13 && age <= 19 {
fmt.Println("Teenager")
} else if age >= 20 && age <= 29 {
fmt.Println("Young adult")
} else if age >= 30 && age <= 39 {
fmt.Println("Adult")
} else {
fmt.Println("Other")
}

Array/Slice

In Python, we are familiar with lists whose size is dynamic. In Go, there are two different concepts for Python lists. The first is an Array, which has a fixed size, and the second is Slices, which is dynamically sized.

Another important thing about Arrays/Slices in Go is that we must define the type of elements we want to store in our Array/Slice. In other words, you can have a slice of strings, not a slice of both strings and integers.

Python:

mix_list = [False, 1, "two"]

Go:

var boolArray [3]bool = [3]bool{false, true, true} // var variableName [array size]type of array elements
var stringArray [3]string = [3]string{"zero", "one", "two"}
var intArray [3]int = [3]int{0, 1, 2}
var boolSlice []bool = []bool{false} // var variableName []type of slice elements
var stringSlice []string = []string{"zero", "one"}
var intSlice []int = []int{0, 1, 2}

For loops Over Arrays/Slices

Python:

mix_list = [False, 1, "two"]
for item in mix_list:
print(item)

Go:

var intSlice []int = []int{0, 1, 2}
for index, value := range intSlice {
fmt.Println(index, value)
}

Maps

Consider Map as the dictionary that we have in Python. Similar to an Array/Slice, you must declare the type of the key and the type of the values.

Python:

the_dictionary = {"hi": 1, "bye": False}
print(the_dictionary["hi"])

Go:

var theMap map[string]int = map[string]int{"hi": 1, "bye": 0}
fmt.Println(theMap["hi"])

HTTP Get Request

Python:

import requests
response = requests.get("https://example.com/")
print(response.content)

Go:

import (
"fmt"
"io"
"log"
"net/http"
)
resp, err := http.Get("https://example.com/")
if err != nil {
log.Println(err)
}
defer resp.Body.Close()
body, err := io.ReadAll(resp.Body)
fmt.Println(string(body))

Conclusion

If you are interested in learning Go, you can check the Go by Example website (I considered its hierarchy of examples for this article) and the Go Tour.

Hope you enjoy coding in Go, and happy learning!

More resources: 

Featured Image: ShadeDesign/Shutterstock

SEO: How to Target Countries

Targeting international consumers in organic search requires a few extra steps. Here are seven tips for higher rankings in the countries of choice.

1. Set up separate sites or sections for each country

To ensure a site ranks in organic search results in more than one country, the options are:

• Create a separate site for each country using country-specific top-level domains (e.g., .de, .uk, .in). Registering all local TLDNs might not be practical, given the cost, restrictions, and unavailability.

• Use a subdomain for each country (e.g., de.yoursite.com). Setting up subdomains requires development and search-optimization resources, such as building links to each subdomain.

• Set up a subdirectory for each country (e.g., yoursite.com/de). Creating directories for each country is often the best approach because the single domain receives all the authority.

2. Use hreflang

Google introduced hreflang in December 2011 to replace its geo-targeting settings inside Webmaster Tools (now Search Console). Hreflang is an HTML attribute — “lang” is language — that helps search engines recognize localized versions of web pages.

3. Translate content to local language

The language of your content is another key indicator to Google as to the country you’re targeting. Google auto-detects language easily. Beyond investing in quality (human) translation, there’s no further action required.

If you are targeting same-language countries, use a region-specific variation for each. A local translator can help.

4. Include location info on the website

Including in the website an address in the country you are targeting helps Google recognize the importance of that locale. This presumes your company has a local physical location or your content addresses one.

5. Attract inbound links from local sites

Google still relies on inbound links to determine the popularity of a website and where to rank it in results, especially when a query has local intent. When targeting another country, launch a local link-building program, perhaps focusing on local journalists.

6. Inform Google Business Profile

If you have a local office, point your Google Business Profile to the regional site.

7. Consider Local Hosting

The location of a web server is a low-level signal to Google, especially for generic top-level domains, such as .com or .net. The larger benefit of hosting near a target market is page speed to those consumers’ browsers.

Some countries, such as Australia, require a license or even a local business presence to host there, making it complicated and expensive. Check local hosting requirements when planning multi-country search optimization.

SEO for SGE

Google’s Search Generative Experience provides AI-generated answers to queries. It’s conversational AI search, in other words, an AI-driven answer engine. Bard, Google’s AI chatbot, powers SGE.

Google launched a limited SGE trial in May 2023 and has since finetuned the layout and responses. A full worldwide rollout is coming, making SGE the future of search.

With that in mind, here are my observations based on extensive SGE testing.

Organic Results

Google’s organic results are not going anywhere, although AI-generated responses:

• Push organic results further down the page,
• Include visual elements such as related images and videos,
• Add links as references for additional info.

SGE’s AI answers are featured snippets on steroids. The answers include “Ask a follow-up,” encouraging searchers to explore the topic further.

Unlike featured snippets, SGE’s answers:

• Include products, maps, and other info based on the perceived search intent,
• Surface user-generated content such as Reddit threads, Quora answers, and customer reviews.

Moreover, SGE answers don’t always answer the query. For example, searching for “best laptops” in SGE includes results such as:

• “RAM vs. Processor — Which Is More Important to Your Business?”
• “What is the importance of a keyboard mouse with a computer?”
• “Computer Basics: Laptop Computers.”

Those results may help a searcher make an informed decision, but they don’t directly respond to the initial query of recommending a laptop.

SGE’s results for “best laptop” include responses unrelated to purchasing. Click image to enlarge.

In June, I reviewed AI-powered search engines. Instead of simply ranking results based on their relevance to a query, AI search provided a wider range of answers. That, I concluded, was the future of search. The latest version of SGE confirms that prediction.

SEO for SGE

SGE uses Google’s existing framework and infrastructure, such as:

• Indexed web pages,
• Knowledge graph, including known entities and associated concepts,
• Shopping results from Google Merchant Center using sellers’ product feeds,
• Map results from Google Business Profile, formerly Google My Business.

Thus optimizing for organic search is largely unchanged:

• Create relevant content addressing the needs of your target customers,

• Optimize product pages and categories based on those users,

• Structure the site to surface pages based on the audience,

• Use Google’s submission tools to ensure content is indexed and thus visible to Bard.

• Obtain external backlinks, as they drive Google Discover and the knowledge graph, among other benefits. Backlinks are especially important for co-citation links in SGE’s results, which place your site next to known entities, thereby making the association.

Google will likely alter SGE many times before the public launch. But SGE’s fundamentals remain and will continue relying on Google’s data. It will encourage clicks to external sites. Despite popular belief, it is not in Google’s interest to keep searchers on its site. Google’s entire revenue model (advertising) relies on clicks.

Organic Search Changes

What will change?

Organic search results will likely become more unpredictable. Thus far, SGE’s answers differ depending on the searcher, making it seemingly impossible to know what folks clicked on to reach a landing page.

Buying journeys will be more complicated. This trend is not new, as Google increasingly serves diverse results that distract searchers from their initial intent. The dynamic nature of AI answers will produce even more extended search journeys.