Members' Insights

A Business Intelligence Approach To Link Building

Cody West
Last Updated: Dec. 13, 2023

Organic search traffic can be a tremendously profitable marketing channel.

To acquire organic search traffic, you need a technically sound website, great content, and high-quality backlinks. Although this may seem like an oversimplification, it’s the foundation for SEO.

Assuming you’ve got your website and content in a comfortable spot, backlinks are going to be the next area you’ll want to address.

Many studies still show backlinks are crucial to ranking your website on Google’s search engine results page. Even Google will tell you that backlinks are one of their algorithms’ top three ranking factors.

A mistake many organizations make is not a lack of investment in link building (although there’s plenty of that). It’s a blind investment in link building.

The economics of a business problem should always be a consideration before execution, NOT after. Organic search traffic acquisition is no different. 

Just because you need backlinks to increase organic traffic, it doesn’t mean the acquisition of links is economical for where you are at as a business.

So how do you determine if it makes sense for your organization to engage in link building?

Unfortunately, there’s no exact science you can follow that produces the same results every time. If that were true, everyone would be doing it. 

However, one of the best ways of increasing your chances of success and reducing risk and uncertainty is by utilizing data to build various models and scenarios that inform your organization’s strategic and tactical decision making. This is called business-intelligence (BI).

If you apply business-intelligence to link building, here are some of the questions you’d want to answer using data:

  • Tactical: How many links do I need to build? What type of links should I build and of what quality? What anchor text should those links contain?
  • Resources: How much is link building going to cost? How long is it going to take?
  • Profitability: Is investing in link building a profitable business decision? How long until I can expect a positive ROI?

When it comes to business-intelligence, there isn’t a one size fits all approach. Every website and industry is different. 

Often, you start with a few questions you want answered. As you dig deeper, new insights will reveal themselves. Your job is to follow the trail the data leaves for you, until you have enough information to choose a link building tool and build a strategy that’s supported by data.

In this post, I’m going to run through the various reports I build that allow me to analyze all the data necessary to design an informed, intelligence-driven link building plan. 

I’ll be using the book writing and publishing company Scribe Media as our example and will be referencing different pieces of an analysis we run at my agency we call our Link Intelligence Report.

You can see a live link intelligence report here and use it to follow along.

Backlink Data Sources

There are many companies that supply backlink data. 

Not only do we want to perform a backlink audit to see our own organization’s backlinks, but we also want to see our competitor’s backlinks (I’ll speak more about this below).

Ahrefs and Majestic have the most comprehensive and useful datasets (in terms of the descriptive statistics they provide), so I’ll be building all the reporting views below using these.

If you’re running an analysis like this on a large site, Google Sheets and Excel are going to be bogged down with the amount of data you’ll be consuming and some of the calculations will be complex and resource intensive, so I recommend doing this with a tool like BigQuery or something similar. 

Breaking Up The Analysis: Domain Level vs Page Level

Pagerank (PR) is an algorithm used by Google Search to rank web pages in their search engine results. PageRank works by counting the number and quality of links to a page to determine a rough estimate of how important the website is. The underlying assumption is that more important websites are likely to receive more links from other websites.”

Wikipedia

According to Gary Illyes, here’s how we know Google still uses it:

Notice how the definition states “web pages” and NOT websites. Google has stated that there is no such thing as overall domain authority as part of their algorithm.

Although domain authority is a made up term used by SEO professionals, what domain authority describes is very real. 

A well architectured website contains internal links on the homepage and other hub pages that point to other pages deeper in the site. If the homepage and other hub pages on the site receive backlinks AND link to other pages on the site, some of the “link equity” (i.e. the value passed to a page from a link) is passed to those pages. This transfer of link equity from page to page to page is what creates the concept of “domain authority”.

If you are trying to rank a single page, it’s almost always better to build links to that page. However, there’s incremental value in building links to other pages as long as there are internal links that end up pointing to the page you want to rank.

When you plan link building for your organization, I recommend splitting the analysis into two parts to account for this:

  • Domain Level: This part accounts for ALL the backlinks pointing at every page on your website
  • Page Level: This part accounts for only the backlinks pointing to the pages you want to rank in Google’s SERPs.

Use Referring Domains, Not Total Backlinks

Throughout this post, I’ll be using the term links or backlinks in place of referring domains. 

You can receive more than 1 backlink from a website. In any type of link analysis, it’s really difficult to account for the diminishing returns of an additional link from the same site, so to compare more fairly, it’s more productive to only analyze the referring domains. 

So how do you pick the most important link from a domain? 

Ahrefs does this for you if you export their backlinks report (1 link per domain) because they take the referring page with the highest URL Rating (UR).

Domain Level Backlink Analysis

The first part of our analysis will be at the domain level.

Let’s dive in.

The Million Dollar Question: How Many Links Do You Need To Close The “Domain Authority” Gap?

Although domain authority isn’t used by Google, the concept it represents is: how much link equity is the website receiving as a whole.

To determine this, you’ll want to measure the “link gap” between your website and your competitors.

First, plot the total number of links to your domain and compare this to your competitors. This approach allows you to map the acquisition velocity of your target website against your closest organic competitors.

The problem with using raw referring domain numbers like above is that not all links are created equal. In fact, there are many links your website will pick up that provide little to no value.

To account for this, create what I call a “Quality Flag” in your dataset that you’ll use to filter out links that don’t meet the criteria you set.

To set your criteria, ask yourself: What descriptive statistics do I use to filter out links I wouldn’t build? What is the minimum threshold for that?

Here’s the criteria I will typically start with:

  • Ahrefs Domain Rating (DR) >= 20
  • Traffic to the linking domain >= 1000
  • Link type = Follow 

Although nofollow is used as a hint to Google, I will usually exclude nofollow links from my analysis because most link builders actively avoid landing these links. If you want to include them, I would set a DR and Traffic threshold on them that’s much larger than what you’d set for follow links.

The idea here is to ONLY include backlinks in your analysis that you would build from a metrics perspective without a manual review of the linking website.

Next, create an additional flag I call Quality Flag 2.0 with requirements that are more strict. Here’s the criteria I will typically set in addition to everything above:

  • Trust flow >= 12
  • Citation flow >= 12
  • Trust flow / citation flow >= 0.8

I will be using the first quality flag for the majority of this analysis. However, Quality Flag 2.0 is useful in the competitor backlink analysis portion of this report.

After applying the Quality Flag to our links, here’s what we’re left with:

Now take the difference between your competitors links and your links to determine how many links you need to build to close the domain authority gap. Break this out by Domain Rating buckets to get a sense of the link quality distribution.

In the table below, a negative number indicates that scribewriting.com has X additional links than the competitor in that DR bucket and a positive number indicates the competitor has X additional links than scribewriting.com.

In this case, you’d want to take the maximum link gap (which is 53 links from self-publishingschool.com) as the number of links you need to close the authority gap.

At this point, you’re not finished because you’d be ignoring a really important concept: link velocity.

According to Ahrefs, “top-ranking pages do tend to acquire more backlinks (and at a faster pace) than the pages that rank below them.” 

That means, each month your competitors are acquiring links naturally at a minimum. They could also be engaging in active link building.

If they acquire links each month at a faster rate than your website, you’ll be fighting an uphill battle. 

To account for link velocity, you will want to calculate the following:

  • The median number of links your competitors built per month over the last 12 months
  • The average number of links your competitors built per month over the last 12 months

To perform this type of competitive intelligence for SEO, first take into account your top 2-4 competitors. Then select the maximum number and then subtract your site’s link velocity. In this case you’d take 16 (the max) and then subtract 8 (your sites link velocity), which leaves you with 8 additional links you’ll need to build each month.

Note: I use both mean and median to get a better sense of the distribution, which can be affected by outliers.

Now, let’s say you want to close the link gap over 12 months. 

Over 12 months, you’ll need to build 96 additional links (i.e. link velocity * 12 months). 

The link gap is 53 links, so you know you need to build a total of 149 links over the next 12 months (i.e. link gap + link velocity over 12 months) or ~12 links per month.

How Much Will It Cost To Build These Links

Now that you have a data backed number for the total number of links to build over the next year, the next question you’ll want to answer is: How much is that going to cost us? 

If you’ve done link building in the past, you should be able to calculate the cost per link.

Here’s the formula:

Link Building Related Expense / Total Links Built

Note: I would only include links in the denominator that meet the Quality Flag that you set.

If you’ve never done link building before, I’d use an arbitrarily high number somewhere in the range of $400 – $600 (although this is HIGHLY dependent on your website and the industry you’re in). 

Now use the following formula to calculate the cost to close the domain authority gap:

Cost To Close Authority GAP Per Month = (Cost Per Link * Total Links Need To Close Gap) / Months To Close Authority Gap

Here’s a look at all of what I just went over in one view:

In-House Teams

If you’ve never ran a link building campaign and plan on doing this in house, I highly recommend running a test. 

First, if you don’t have an experienced digital PR professional on your team, I would purchase a training like Launchpad PR that provides you with all the SOPs and systems you’ll need to run your own link building campaigns. It will dramatically speed up the time it takes to close the knowledge and experience gap, so you’re building links effectively.

Once your team is brought up to speed, run a test over a 3 month period to see how many links your team can build. After the 3 months is up, calculate your cost per link.

You can decrease your final cost per link by a factor of 10-20%, which would be assuming your team gets more efficient over time.

If this sounds too complicated, you should hire an experienced link building agency. They will be able to give you a fairly accurate cost per link number to use in the calculations above.

Agency Teams

If you’re an agency, you should have a decent idea of how much your cost per link is and you should be doing these calculations before engaging with a client to ensure value add.

If you’re an agency looking to add link building as a service, I recommend following what I mentioned in the in-house team section above. Remember, cost-per-link it will vary across industries.

Tactical Link Profile Analysis

Now that you have a ballpark cost for link building, you’re going to want to go deeper into your competitor’s link profiles to get an understanding of the strategies and tactics they’re using to build links.

First, determine the homepage versus inner page link distribution:

Notice how Scribe Media’s home versus inner page distribution is flipped compared to their competitors. If you dig into this, you’ll find that Scribe Media has a few high-profile co-founders that do very effective PR, but do NOT do any form of deliberate link building. PR links usually always point to the homepage with branded anchor text, so this distribution makes sense.

The takeaway for Scribe Media here is to engage in link building to their inner pages (i.e. blog posts).

Second, calculate the percentage of pages on each website that have 1 or more backlinks that meet our Quality Flag:

Higher percentages indicate more equal distribution of link equity throughout the website. This is a good thing and means that with a good internal linking strategy, the site will be able to effectively distribute the link equity throughout the website.

Scribe Media’s distribution here is low due to the homepage receiving the majority of links (from PR) and their lack of link building efforts.

Third, we’ll want to calculate the types of backlinks that Scribe Media’s competitors have.

Understanding the link type distribution gives you insights into the link building tactics you’ll want to engage in. 

For Scribe Media, the majority of links are content-based links, so guest posting and outreach campaigns should likely be the bulk of their link building efforts. Resource pages are also a big portion of the total links, so Scribe Media should look at a handful of the top resource pages and consider creating content to pitch to the resource page owner.

Next, you’re going to want to know what the main topic(s) of the websites that are linking to your competitors.

Topically relevant backlinks are preferred. Without getting into the algorithm itself, just think about it: if you are a book writing and publishing site, would you prefer a backlink from a writing blog or a golden retriever blog?

The writing blog is MUCH more likely to refer customers to you than the golden retriever blog, so you pick the former. 

Majestic has a descriptive statistic called topical trust flow that is useful in identifying the main topic(s) of a website.

Create a reporting view that shows topical trust flow distribution by competitor:

No surprise here…

The largest topical trust flow category is Arts / Writers Resources followed by Reference / Education, and Arts / Literature respectively. 

When we start deciding which sites to reach out to, this piece of information will be incredibly useful. Another interesting thing to point out here is that 19% of self-publishingschool.com’s links are Reference/Education, which is the largest percentage present. 

I would guess that they either have some really solid resources and are receiving resource page links OR they have a scholarship page and are engaging in scholarship link building.

Next, you will want to look at what individual webpages on your competitor’s websites acquire the most high-quality backlinks. If you’ve ever seen Ahref’s Best By Link report, you’ll want to build something similar. 

The difference is you’ll want to include ALL of your competitors and filters out any links that don’t meet your Quality Flag. In the visualization below, I’m calculating the mean, median, and sum of URL Rating (UR) to get a sense of the strength of the links pointing to each page.

If you look at the best URL, it’s a scholarship page on self-publishingschool.com. This explains the 19% of links that have a topical trust flow of Reference / Education above. When you select pages to build links to, you’ll be able to use the visualization above to help select the most “linkable” pages based on your competitor’s content.

Next, check the follow versus nofollow distribution across the whole site.

If you’re not actively building links, you don’t need to worry about this, but it’s a good health check to make for your website. This distribution is much more important at the page level, which we’ll cover below.

In some industries, you can build a ton of really good links to images. You’ll want to see what percentage of links pointing to your competitors websites are image links.

After filtering out links that don’t meet our Quality Flag, we come to the conclusion that image links aren’t going to be important in the book writing and publishing industry.

Hacking The Link Prospecting Process

Now that you have a good overview of the link building landscape from a domain level, you’ll want to build a competitor backlink analysis to highlight the individual links your competitors have that you do NOT have.

In my opinion, your competitor’s backlinks are the best place to start when it comes to link prospecting and uncovering sites you want to link to you.

Here are the questions you’ll want answered:

  • How many competitors have a link from an individual website?
  • What links do your competitors have but you don’t?
  • What is the link category and topic of the linking domain?
  • What are the SEO metrics for each link, so you can prioritize the best links and filter out the junk.

Here’s what you’ll want it to look like:

The above view is filtered so it’s only showing links that Scribe’s competitors have that, but that Scribe does not have.

Let’s say Scribe want to run a guest posting campaign because they know they need to build more content-based links. 

Use the filters to filter for all domains that have a Topical Trust Flow of Arts/Writers Resources (because that’s the most prevalent topic based on my previous learning), a DR of 50 or above, and that are either content-based links or guest posts. 

Then you can sort your list by DR so you see the stronger websites first.

Now you can simply export this list, find contact information for each site, and then reach out to them regarding a guest post.

In my opinion, this is the best way to start link prospecting.

Note: Here’s a simpler version of a competitor backlink analysis I created in Google Sheets.

Page Level Backlink Analysis

Now that our domain level analysis is complete, we can start on the page level analysis.

Like I mentioned above, page rank is calculated at the page level, NOT the domain level. For a page level analysis, I recommend selecting the 5 – 10 pages on your website that are most important from a business perspective OR that you know you need to build a significant number of links to (use Ahrefs Keyword Difficulty to do this).

At Assisted Reach, we do an analysis we call a “URL Prioritization” that takes into account factors like competition, traffic potential, business value, and more to prioritize the pages on a website by importance.

However, you probably already know the top 5 – 10 pages off- hand that you want to rank for, so I would just go with those.

For each page, you’ll want to determine what the main keyword is and then pull data for the top 10 pages ranking for the main keyword.

Note: There’s no point in doing this for more than 20 or so pages at a given time. When engaging in link building, you’re not going to be targeting more than a few pages at a time for efficiency reasons. Once you rank your first few pages, perform the analysis again on the next few pages. 

SERP Overview

The first view you’ll want to create is the number of links that meet your Quality Flag by Domain Rating (DR) bucket for each keyword. Include page type, page category, and rank for each URL. 

Here’s an example for the keyword “book cover design”:

Next, calculate the average and median for each column and then take the max of the average and median (e.g. if the avg is 4 and the median is 5, the max would be 5). 

Subtract the links your page has from the maximums you calculated to get the “link gap”. The link GAP is the number of links you should plan on building to rank for the keyword. 

Here’s what it will look like once you’re finished:

The table above states that Scribe Media should build ~22 total links. 

  • 2 of those links between DR 20-29 links
  • 4 of those links between DR 30-39 links
  • 4 of those links between DR 40-49 links
  • 7 of those links between DR 50-59 links
  • 4 of those links between DR 60-69 links
  • 5 of those links between DR 70-79 links
  • 3 of those links between DR 80-89 links
  • 2 of those links between DR 90-100 links

You should really be building the best links you can, but use this as a gauge for your minimum target for link prospecting purposes.

If you’re a site with a really high domain authority (or DR) compared to the other sites in the SERP you’re competing in, you may want to take the minimum rather than the max because it will be easier for you to rank.

You also may want to use a single competitor as your benchmark, rather than using average and median logic. It’s all situationally dependent. 

When I approach link building, I prefer to overestimate what I need. For example, I’d rather estimate I need 25 links when I really need 15, rather than the inverse. The reason has everything to do with cost forecasting.

Link Velocity At The SERP Level

Just like what we did at the domain level to account for link velocity, we’ll do the same at the page level.

First, plot the links that meet your Quality Flag for each competitor to visualize link velocity.

Next, calculate the mean and median over the last 3 months and look at the distribution by competitor.

Note: Again, I use both mean and median to get a better sense of the distribution, which can be affected by outliers.

Typically, I’ll take the maximum total (either mean or median) and use that as my link velocity number, but I’ll also look at it by competitor; just in case I see someone who’s doing a really good job at actively building links. 

For example, I know that Canva is actively building backlinks. TTT has a Slack channel dedicated to link building and I’ve seen screenshots of great outreach emails from Canvas talented SEO team.

In this case, Scribe Media will likely never outrank the top 3 sites because of the content types (lead gen pages versus blog posts), so we’ll use 1 as our link velocity number.

Note: When you see various content types in a SERP, that generally means there is varying intent for that single keyword.

Now let’s say we want to close the link gap in 6 months. 

Over those 6 months, the competition will build an additional 6 links:

Link velocity of 1 per month * 6 months = 6 backlinks

We also need 22 links to close the link gap, so we’ll need to build 28 total links:

22 links to close the gap + 6 links to account for link velocity = 28 total links needed

How Much Will It Cost To Build These Links

Now that we know how many links we want to build to each page, calculate your historical cost per link using this formula:

Cost Per Link = Historical Link Building Related Expense / Total Historical Links Built

Again, only include links in the denominator that meet the Quality Flag that you set.

If you’ve never done link building before:

  • Use an arbitrarily high number somewhere in the range of $400 – $600 OR 
  • run a ~3 month test and after the test is over, calculate your cost per link assuming a small (e.g. 10-20%) efficiency gain as your people and process improve. 

Now that you have your cost per link, repeat this for each keyword/page you’re analyzing. Then, use the following formula to calculate the total cost in links to rank for all those keywords/pages.

Total cost in links = Total links needed * Cost per link

Now you should know:

  • How many links you need to rank for your selected keywords/pages
  • The cost in links to rank for your selected keywords/pages

We’ll be coming back to these two numbers at the end of this post when we develop our final link building plan and calculate ROI.

Anchor Text

Anchor text is incredibly important when it comes to link building. I’m not going to go into too much detail here because there are already some excellent guides written on the subject.

Here’s one and here’s another

The anchor text questions you’ll want to answer are the following:

  • What is my target anchor text distribution for each keyword?
  • What anchors should I use?
  • What anchor do I use for each link category?

Anchor Text Distribution

First, calculate the anchor text distribution for the web pages ranking in the top 10 for your keyword:

Note: The “Client Anchor Type Distribution” contains dummy data in this example, not actuals.

Here are the definitions I use (these definitions make automating this logic simple). You could do this manually if you want to break the anchor text up into different categories.

  • Seed: The exact keyword. If the keyword is “writing tips” the only anchor text that would count as seed is “writing tips”
  • Exact: The anchor text contains “writing tips” in addition to other words either before, after, or both
  • Partial: The anchor text contains “writing” or “tips” or both, but not in the same order
  • URL: The anchor text is a url
  • Other: This is anything else that doesn’t meet the above criteria.

I always prefer being risk averse when it comes to anchor text. That means avoiding an anchor type distribution that is too heavy in the Seed or Exact anchor type.

For example, let’s say we want to build the following anchor text profile:

  • 1 seed
  • 1 Partial
  • 2 Exact
  • 4 Other
  • 2 URL

Instead of building the Seed and Exact anchor first, build a couple URL and Other anchors before building the Seed anchor. Next, build 1 Partial and 1 Other before building an Exact. 

There’s no hard and fast rule here and it’s total speculation. I would just make sure that whatever you do, it’s something that would naturally occur in the online world.

Tip: Because you can control the anchor text when guest posting, choose Exact and Seed anchors for your best links if you can.

Anchor Text Ideation

Once you know the anchor type distribution you’re shooting for, you’ll want to know what anchors you should use. The easiest way to do this is to pull in your competitors anchor text.

Viewing your competitors anchors is great for a number of reasons, but my two favorite are:

  1. It gives you a good idea of the different contexts your page can be linked for
  2. It allows you to plan out your anchors without having to spend time brainstorming

As mentioned before, guest posting is one of the only link building tactics where you have control over the anchor text. 

The above view is especially useful for guest posting because (as mentioned before) guest posting is one of the only link building tactics where you have control over the anchor text.

Choosing The Right Anchor For Each Link Type

I’d also recommend you look at the anchor type to link category distribution: 

This view will allow you to choose the correct anchor text depending on the link building strategy you are employing.

Hacking Link Prospecting… Again

To speed up the link prospecting process, build the same competitor backlink analysis as you did at the domain level, but use the page level data instead:

Bringing This All Together To Build Out A Strategy & Determine ROI

After this analysis, we have an incredibly good idea of what’s working for our competitors from a link building perspective, how many links we need to build, where those links need to point to, and more.

The last step here is to calculate a return on link building and then finalize our strategy and tactics.

Using A Monte Carlo Simulation To Forecast Link Building ROI

Using the information you glean from this analysis, you can use what’s called a Monte Carlo simulation to forecast the return on link building. 

I wrote a post you can find here that takes you through step-by-step how to calculate link building ROI using a Monte Carlo simulation.

Strategy / Tactics Summary

We know we need to build ~149 links to close the DA gap over 12 months. We’re going to build 85 of these 132 links to the 6 pages/keywords we analyzed. The remaining we’ll wait to see our results before making the call.

From a tactical perspective, we know we want these links to point to inner pages and that the majority of links we’ll build are guest posts and resource pages. To ensure the links are topically relevant, we’ll look for a TTF of Arts/Writers Resources, and then may want to consider creating a scholarship.

Finally, we’ll use both competitor backlink analyses to speed up the link prospecting process and then begin our link building campaign assuming all the numbers make sense from a return perspective.

Link and Anchor Text Tracking

Now that you know all this, the last piece is to track the links you begin building. For tracking, I like to manual track the following data points for each link I build:

  • Keyword
  • Target URL
  • Linking URL
  • Domain Rating
  • Traffic To The Linking Domain
  • Link Category
  • Type (follow vs nofollow)
  • Anchor
  • Anchor Type
  • Anchor Use Count ( so you aren’t using the same anchor text more than a couple time)
  • Date Link Was Built
  • New Link Flag (this indicated whether it was a link you had prior to the analysis versus a link you built after the analysis)

Here’s a dummy Google Sheets template you can use for tracking purposes. The RED cells indicate there’s an ARRAYFORMULA, so you don’t want to delete these.

Once I’ve got my tracking template set up, I add all the links I already have to it for the 5 – 10 pages I’m planning on building links to and then build the following views to track my link building progress and anchor text profiles.

Final Thoughts

One of my favorite quotes about analytics is one from Douglas Hubbard where we says:

Anything can be measured. If something can be observed in a way at all, it lends itself to some type of measurement method. No matter how fuzzy the measurement is, it’s still a measurement if it tell you more than you knew before”

Douglas Hubbard

Link building is no different.

Through the use of the amazing datasets available to you combined with business intelligence, you can reduce uncertainty and make better, data-driven decisions when planning link building for your organization.