We compiled a collection of SEO statistics a few months ago.
Although it appears to be a fairly typical piece, we deliberately crafted it to draw links from an outreach campaign.
And it succeeded. We sent 515 emails, and 32 websites responded with 36 backlinks.
The article is currently ranked number one for “SEO stats”:
In this guide, you’ll learn:
Why we chose a statistics page
How we created the page and built links to it
Detailed results from our campaign
How we could have improved our results
Why did we create a statistics page?
Most likely, you are familiar with the skyscraper technique. It’s a link-building strategy in which you locate a page with lots of links, produce something “better,” and then approach the people who are linking to the now inferior page with your new and improved resource.
It can work for any type of content, but we had a theory that it might work particularly well for statistics pages because:
Statistics pages tend to have a lot of links
Statistics pages are often outdated
Statistics are often sloppily updated
Let’s delve a bit deeper into why each of these things matter.
1. Statistics pages tend to have a lot of links
Each “skyscraper” campaign begins with one or more pages that are loaded with links. And if you search Google for statistics pages about any popular topic, you’ll almost always see that the top-ranking pages have tons of backlinks.
Just have a peek at the “youtube statistics” search results:
This site has numerous pages with links to tens of thousands of websites.
This frequently occurs as a result of the frequent use of these pages’ statistics by bloggers and journalists in their writing.
For example, look at this post from Shopify about starting a YouTube channel for your business:
The writer cites two statistics in the first paragraph and links to his sources, one of which is a curated list of YouTube statistics. It’s highly likely that he came across this post while doing research for his article, found a useful statistic, then cited and linked to the source.
2. Statistics pages are often outdated
Because most statistics pages are rarely updated, they frequently contain stale and incorrect data.
Here’s only one illustration:
If we click on the link for that statistic’s source, we are directed to YouTube’s Press page. Instead of the 1.9 billion users that are cited, it says that there are over two billion monthly YouTube subscribers.
This is problematic because it causes bloggers and journalists to use outmoded data.
For instance, if we examine the Backlinks report for this page in Site Explorer, look for the word “1.9 billion” in the link anchors and surrounding text, then enable the “one link per domain” button, we find 55 websites referring to the page and using the out-of-date statistic.
3. Statistics pages are often sloppily updated
Even when users update their statistics sites, they frequently make mistakes. They frequently make the error of deleting out-of-date statistics from the page without replacing them with newer ones.
The issue here is that the citations and connections continue even after the numbers are removed.
Check the Anchors report in Site Explorer, for instance, to see a list of YouTube statistics. Numerous referring pages make reference to the fact that YouTube is the third most popular website on the internet. But when we look for that statistic on the page itself, we can’t find it.
That’s because the author deleted that statistic after updating the page.
What we did
You may have a general notion of where we’re going with this already. Don’t worry if that’s not the case. The complete procedure will be detailed here.
Here’s what we did:
Found a winning topic
Found link prospects
Narrowed down the link prospects
Created the statistics page
Found contact information and vetted prospects
Wrote and sent outreach emails
Step 1. Find a winning topic
Due to the focus of our website, we knew we needed to make a statistics page on SEO or internet marketing. However, we still needed to do some research because we didn’t yet have a certain theme in mind.
We began by looking for keywords relating to our industry in Ahrefs’ Keywords Explorer, such as SEO, search engine optimization, content marketing, etc.
Then, using the “Include” filter in the Phrase Match report, we looked for keywords that contained phrases like statistics, stats, facts, and figures.
When doing this, be careful to switch the “Include” filter’s settings from “All” to “Any.”
This provided us with a ton of suggestions, but we needed to focus on those that had high-ranking pages and numerous backlinks. To do this, we filtered out all low-difficulty keywords using the Keyword Difficulty filter.
If you’re asking why it’s because the statistic is based on the typical number of referring domains to the current top-ranking pages. A high KD score for a term typically indicates that many different websites are linking to the top-ranking pages.
This significantly reduced the options, but a few remained, with “SEO data” standing out as the most pertinent.
In order to confirm that the top-ranking pages were primarily well-picked statistics pages, we lastly evaluated the SERP overview. The page titles made it quite simple to determine whether this was the case. If you want to learn about the Link Bait, Click here
Step 2. Find link prospects
Finding link candidates would be easy if we were conducting a “shotgun” campaign in this case. Then, after writing any old piece and sending the identical outreach email to everyone, we would simply download the entire list of backlinks to each competitive statistics page.
This would most likely say, “Hey, see you connected to this statistic page. Ours is superior. Perhaps change the link?”
This is the traditional “Skyscraper” strategy, and we don’t like it. All of us have already received emails like this one from outreach campaigns. They are unpleasant, useless, and spammy.
So, this is what we did in its place:
First, we returned to the “SEO statistics” SERP overview and searched for the curated stats page with the greatest number of backlinks.
Second, we looked at this page’s analytics to determine which ones generated the most backlinks.
To achieve this, we accessed Site Explorer’s Anchors report for the page and searched for stats that were frequently mentioned. We immediately noticed a large number of sites mentioning how 93% of internet experiences start with a search engine.
After entering this information into a spreadsheet, we perused the report further until we found a few well-liked statistics.
Next, we wanted to tick off two checkboxes for each statistic:
Are there enough people linking because of this statistic?
Can we justify a good reason for them to link to us instead of, or in addition to the current page?
Initially, we toggled the “one link per domain” filter in the Backlinks report for the page, then we searched for each statistic in the link anchors and surrounding words.
Over 700 websites were connecting to the page with the “93%” number.
The first box has been checked.
Then, we entered the stats page and performed the following steps to determine whether there was a valid cause to get in touch with those people:
We discovered that the “93%” statistic was not even listed on the page.
When we checked their sources, we found that many of them were antiquated. We searched the web for a current statistic wherever this was the case.
Everything was then recorded in a spreadsheet to give us something that looked like this:
Column A: The URL of the statistics page (where we found the statistic).
Column B: The number of linking referring domains citing the statistic (we also linked this to the filtered report in Site Explorer for easy access).
Column C: The statistic itself.
Column D: The original source of the statistic (if we could find it).
Column E: The age of the statistic.
Column F: The newer statistic (where applicable).
Column G: The source of the new statistic.
Column H: The age of the new statistic.
Column I: Notes and pitch angle ideas, such as whether the statistic was outdated or missing from the page.
Once we had thousands of prospects based on eight statistics, we then repeated the entire process for the remaining top-ranking statistics pages.
Then, we imported them into Google Sheets, tagged each URL with its segment, and collected all of the pertinent prospects for each statistic and page from the Backlinks report.
We had 1,986 unchecked URLs in all.
Step 3. Narrowing down link prospects
We had a cause to contact 1,900+ prospects, but for obvious reasons, we didn’t want to contact low-quality websites. Thus, organizing our list of potential customers was our next step.
Since websites may connect to multiple statistics pages, the first step was to remove duplicate URLs from the same domain. Additionally, we didn’t want to contact the same website repeatedly.
Doing this wasn’t that difficult. We simply created a column for the referring website and extracted the root domains in bulk using a program similar to this one. After that, we adhered them to our sheet.
After that, we eliminated duplicate domains using the built-in capability.
The next step was to delete prospects who were using sponsored, nofollow, or user-generated content links to link to the analytics pages. We deleted the rows by filtering the “Type” column to achieve this.
Excluding sites with little traffic was the last step. We used Google Sheets’ script editor and the Ahrefs API to gather domain-level traffic for all prospects.
Use our Batch Analysis tool if you’re following along but don’t have an API subscription. Simply paste up to 200 domains at once and choose domain with all its subdomains as the target mode.
After that, you can run a VLOOKUP on the root domains and export the file.
We have 902 prospects left on our sheet after this procedure.
Step 4. Create the statistics page
The reality of statistics pages is that it’s tough to make one that stands out from the crowd. However, we did three main things to make ours the “best” SEO statistics page:
Included popular statistics from other pages
Included other interesting statistics
Grouped statistics by category
Let’s go through why these things are important.
Including popular statistics from other pages
Each statistic that assisted other pages earning plenty of links identical to theirs was incorporated in our post. We identified a more recent statistic if the original one was out of date and included it.
We did this, in part, to link our page to our link-building strategy. We also observed that people regarded these figures to be the most helpful, though. How are we aware? Because bloggers and journalists commonly reference them in their posts.
Including other interesting statistics
Not all of the statistics on our page related to our link-building activity. We searched for and included others that we thought could be of interest.
However, we took an attempt to track down the original source for each statistic we wanted to include because we didn’t want to include any stale data. We looked for an updated source if we couldn’t discover the original or realized it was out of date.
Although it may seem insignificant, most statistical lists have a serious problem with this.
For instance, the “93%” figure we previously discovered is from a 2006 publication.
Grouped stats by category
Statistical tables are not as well-read as a typical blog post. Most are merely looking for carefully chosen data to include in their writings to bolster their assertions. We have to structure and make our post simple to read because of this.
To do this, we added jump links to the introduction and organized the information into categories.
Additionally, we included a list of the top SEO statistics at the beginning of the post under the heading “top SEO statistics.” Finding the most interesting and often referenced statistics became simpler as a result.
Step 5. Find contact information and vet prospects
After publishing our SEO data list (and de-indexing the page), we were almost ready to engage in some outreach and link-building. But first, we had to track out our prospects’ contact information.
If you have read our manual on link creation at scale or watched our film on the subject, you are aware of our preference for leveraging APIs and automation in this situation.
With that in mind, here’s what we did:
We ran all the URLs through a custom tool to scrape as many author names as possible. It was far from perfect, but we ended up with 741 names nonetheless.
We took those names and ran them through Hunter’s API to search for an email address. Hunter sent back 452 email addresses. That’s around a 60% hit rate.
We ran these email addresses through NeverBounce’s API to see which of them were deliverable.
We had 168 valid emails after around 30 minutes of automation, 92 catchalls that required some manual assistance, and 178 emails that were not deliverable.
It was simple to determine whether an email was legitimate by examining the pages’ general condition. One of our team members, Vlad, managed this operation. He put in a few hours per week screening potential clients and tracking out missing contact details.
Here is a condensed description of the screening procedure he underwent:
The spreadsheet was then updated with his tags, leaving us with 515 sites that were prepared for a pitch.
He also noted the subject matter of the pages that linked to them.
Step 6. Write and send the outreach emails
After around 30 minutes of automation, we had 168 deliverable emails, 92 catchalls that need some manual work, and 178 that were not.
Our only requirement for legitimate emails was that the pages be of a reasonable standard. Vlad, a member of our team, managed this operation. Each week, he put in a few hours to research potential clients and track down any information that was lacking.
The screening procedure he underwent is summarized in the following manner:
After that, he annotated each prospect in the spreadsheet, leaving us with 515 sites that were prepared for a pitch.
He also noted the subjects of the pages that were linked.
We sent 515 emails in total. 42 of the 473 deliveries bounced.
Our conversion rate, calculated based on deliverable emails, is 5.71%, which means that 27 of the websites we contacted included links to us.
However, two other cool things happened:
We got links from five sites that we didn’t contact. This is likely because some people found our post from others that linked to us and from the odd social shares. In total, we had 19 shares on Facebook and two on Twitter.
Some websites linked more than once. This was from both new and old pages.
After accounting for those factors, our campaign netted us 36 editorial links in total from 32 different websites.
Of course, quantity alone cannot win the race. How about the standard?
Let’s categorize everything by Domain Rating:
A DR of 70 or more was present in 9 of the 32 referring domains. 12 had DRs between 40 and 69. The 11 others had DR values ranging from 4 to 39.
Here is what we get if we do the same with anticipated organic traffic:
Over 1,000,000 search visits per month are made to one website. 6 receives between 10,000 and one million. 6 win anywhere from 1,000 to 9,999. 16 score somewhere from 100 to 999. 4 also receives no search traffic.
Therefore, based on traditional SEO measures, the majority of the links we received were of good quality. Although these metrics aren’t infallible, I would concur with them after manually checking each link we acquired.
Could we improve these results?
Despite the campaign’s success, some could contend that a conversion rate of 5.71% is not particularly noteworthy.
That may have been the case five years ago, but outreach is becoming more difficult, and more people are requesting money or other rewards.
Having said that, there are two ways we probably could have raised our conversion rate.
1. Negotiate with prospects
Our campaign had a conversion rate of 5.71%, while the reply rate was 17.55%. That means that although 83 people replied to our email, just 27 of them linked to us (plus the two we didn’t contact).
Many of these were requests for link exchanges and other things, however some of them were “thanks, but no thanks” responses.
Here’s a quick breakdown:
8 people requested a link exchange;
6 people requested something else in return (e.g., free Ahrefs account, to share their content on social media, etc.);
3 people asked for money
Purchasing links is prohibited by Google’s Webmaster Guidelines, thus we would never engage in it or advocate for others to do so. However, even if we disregard them, there were still 14 individuals who were open to compromise. We probably could have reached a mutually advantageous agreement and persuaded these people to link to us with a little back and forth. By doing so, we would have reached 41 referring domains, representing an 8.7% conversion rate.
2. Send follow-ups
Sending three follow-up emails increased their results by at least double, according to Authority Hacker’s analysis of over 600,000 outreach emails. But because we didn’t want to annoy folks, we didn’t send any automatic follow-ups.
Most link builders will think that’s nuts, but we simply launched one campaign to see how it worked. Not obtaining as many links as we could was not our aim. Because of this, we only followed up with those who had promised to link to us but hadn’t done so in the previous two weeks.
Our link acquisition rate would have been at least 11.42% if we had followed Authority Hacker’s recommendation and issued three follow-ups.
You could wonder if blogger outreach is still worthwhile given that we put a lot of work into this campaign and only ended up with backlinks from 32 websites. Fortunately, the response to this query is uncomplicated: of course it is.
Backlinks are still a significant ranking factor, and outreach is the only way to build high-quality, white-hat links. It’s important to note that even though we ran our campaign across a few weeks, setting it up wasn’t too difficult. We believe that we could build up the full campaign in one working day if we were to try it again, presuming that we had someone screening candidates and locating contact information.